Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Investigations in x-ray computed tomography polyacrylamide gel dosimetry Hilts, Michelle Louise 2005

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2005-104023.pdf [ 18.21MB ]
Metadata
JSON: 831-1.0085737.json
JSON-LD: 831-1.0085737-ld.json
RDF/XML (Pretty): 831-1.0085737-rdf.xml
RDF/JSON: 831-1.0085737-rdf.json
Turtle: 831-1.0085737-turtle.txt
N-Triples: 831-1.0085737-rdf-ntriples.txt
Original Record: 831-1.0085737-source.json
Full Text
831-1.0085737-fulltext.txt
Citation
831-1.0085737.ris

Full Text

I N V E S T I G A T I O N S IN X - R A Y C O M P U T E D T O M O G R A P H Y P O L Y A C R Y L A M I D E G E L D O S I M E T R Y by M i c h e l l e L o u i s e H i l t s M . S c , U n i v e r s i t y of B r i t i s h C o l u m b i a , 1999 B . S c , M c M a s t e r Un ivers i ty , 1996 B . A . , M c M a s t e r U n i v e r s i t y , 1996 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D o c t o r o f P h i l o s o p h y i n T H E F A C U L T Y O F G R A D U A T E S T U D I E S (Physics ) T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A M a y 2005 © M i c h e l l e Lou ise H i l t s , 2005 Abstract P o l y a c r y l a m i d e gels ( P A G s ) are radiosensi t ive mater ia l s cur rent ly under develop-ment for use as three d i m e n s i o n a l (3D) dosimeters i n r a d i a t i o n therapy. Dose i n -f o r m a t i o n is recorded i n the gels a n d e x t r a c t e d t h r o u g h i m a g i n g . X - r a y c o m p u t e d t o m o g r a p h y ( C T ) has emerged as a p r o m i s i n g gel i m a g i n g m e t h o d due to a change i n gel dens i ty that occurs u p o n i r r a d i a t i o n . T h e access ib i l i ty of C T technology to cancer hospi ta ls makes C T read-out c l i n i c a l l y a t t rac t i ve , however the technique remains of l i m i t e d c l i n i c a l use due i n p a r t to poor dose reso lut i on . T h i s thesis inves-t igates the use of C T for e x t r a c t i n g dose i n f o r m a t i o n f r o m P A G w i t h a n overa l l goal to improve achievable dose reso lut ion . Thes i s results are d i v i d e d in to three studies: a gel c ompos i t i ona l s tudy, a s t u d y of noise a n d dose reso lut i on a n d a d i g i t a l f i l t e r ing study. T h e first s t u d y investigates the effects of gel c o m p o s i t i o n o n P A G C T dose response a n d the u n d e r l y i n g dens i ty change. Sys tems for i r r a d i a t i n g a n d i m a g i n g gels are designed a n d tested a n d dose response r e p r o d u c i b i l i t y is establ ished. R e -sults ind i cate d r a m a t i c v a r i a t i o n i n C T dose response sens i t iv i ty a n d range w i t h gel c ompos i t i on . A m o d e l is developed to describe gel dens i ty change w i t h dose, revea l ing two fundamenta l propert ies of the dens i ty to dose response: the dens i ty change t h a t occurs per u n i t p o l y m e r y i e l d is highest for gels w i t h low a n d h i g h concentrat ions of c ross l ink ing molecule (%C) a n d the dose response sens i t i v i ty is l i n e a r l y dependent on the t o t a l c oncentra t i on of m o n o m e r i n the gel. T h e second s t u d y investigates strategies for m i n i m i z i n g noise i n x - r a y C T p o l y m e r gel dos imetry a n d assesses sys tem per formance . Speci f ical ly , the effects of p h a n t o m design, s cann ing technique a n d image voxel size o n image noise are invest igated. T h i s w o r k leads to the establ i shment of a m e t h o d of p r e d i c t i n g image noise for any g iven C T i m a g i n g pro toco l . Image u n i f o r m i t y is also assessed, i n the context of noise levels i n gel dos imetry . T h e effect of s cann ing p r o t o c o l o n i m a g i n g t i m e is establ ished a n d the dose reso lu t i on achievable w i t h a n o p t i m i z e d sys tem is ca l cu la ted g iven voxe l size a n d i m a g i n g t i m e constraints . These results , w h e n c o m p a r e d w i t h p u b l i s h e d values for M R I a n d o p t i c a l C T gel d o s i m e t r y ind i ca te t h a t C T dose reso lut ion (e.g. 5%, 1 x 1 x 3 m m 3 voxels) , is s t i l l not at the level of the best M R I or o p t i c a l C T techniques, however fast i m a g i n g t imes makes the r a p i d acqu i s i t i on of v o l u m e t r i c d a t a most feasible w i t h x - r a y C T . T h e t h i r d s t u d y investigates the p o t e n t i a l of image f i l t er ing for i m p r o v e d dose reso lut ion i n C T gel dos imetry . C T image noise is charac ter i zed as G a u s s i a n d i s t r i b u t e d a n d independent of s igna l s t r e n g t h a n d niters for r e d u c i n g s p a t i a l l y invar iant noise are invest igated : mean , m e d i a n , m i d p o i n t , adapt ive m e a n , a l p h a -t r i m m e d m e a n , s i g m a m e a n a n d a r e la t i ve ly new fi lter ca l led S U S A N . T h e fi lters are tested on a C T image of a P A G i r r a d i a t e d w i t h a c l i n i c a l l y relevant dose d i s t r i b u t i o n . F i l t e r per formance varies great ly i n b o t h achieved dose reso lut i on a n d affects o n the s p a t i a l d i s t r i b u t i o n of dose. T h e A D A P T I V E a n d S U S A N filters prov ide the best overa l l per formance , more t h a n h a l v i n g the dose reso lut ion w i t h o u t s ign i f i cant ly d i s t o r t i n g the s p a t i a l d i s t r i b u t i o n of dose. I n summary , th i s thesis provides new ins ight in to the f u n d a m e n t a l n a t u r e of P A G densi ty to dose response, develops strategies for m i n i m i z i n g image noise a n d quantif ies sys tem per formance a n d demonstrates that d i g i t a l image f i l t e r ing is a n effective t o o l to prov ide a d d i t i o n a l improvements to dose reso lut ion . Contents A b s t r a c t i i L i s t o f Tables ix L i s t of F i g u r e s x i L i s t o f A b b r e v i a t i o n s x x A c k n o w l e d g e m e n t s x x i i C h a p t e r 1 I n t r o d u c t i o n 1 1.1 R a d i a t i o n T h e r a p y 2 1.1.1 O v e r v i e w 2 1.1.2 M o d e r n Techniques 6 1.2 R a d i a t i o n D o s i m e t r y Bas i c s 11 1.2.1 A b s o r b e d Dose 11 1.2.2 Interact ions of R a d i a t i o n w i t h M a t t e r 12 1.3 Dos imeters i n R a d i a t i o n T h e r a p y 17 1.3.1 Dos imeter Requirements 17 1.3.2 C o m m o n Dos imeters U s e d i n R a d i a t i o n T h e r a p y 19 1.3.3 T h r e e D i m e n s i o n a l D o s i m e t r y 24 1.4 Thes i s G o a l s 29 C h a p t e r 2 X - R a y C o m p u t e d T o m o g r a p h y 31 2.1 I n t r o d u c t i o n 32 2.1.1 A B r i e f H i s t o r y of C T 32 2.1.2 Genera t i ons of C T Scanners 34 2.1.3 F u n d a m e n t a l P r i n c i p l e s of C T 37 2.2 M o d e r n C T Scanners 40 2.2.1 S y s t e m O v e r v i e w 40 2.2.2 M a j o r C o m p o n e n t s 41 2.2.3 Technique P a r a m e t e r s 45 2.2.4 Recent A d v a n c e s 49 2.3 Image R e c o n s t r u c t i o n 50 2.3.1 I terat ive R e c o n s t r u c t i o n 51 2.3.2 B a c k p r o j e c t i o n 52 2.3.3 F i l t e r e d B a c k p r o j e c t i o n 53 2.4 Noise a n d A r t e f a c t s 56 2.4.1 Noise 56 2.4.2 A r t e f a c t s 57 C h a p t e r 3 P o l y m e r G e l D o s i m e t r y 59 3.1 I n t r o d u c t i o n 59 3.1.1 A B r i e f H i s t o r y of G e l D o s i m e t r y 59 3.1.2 O v e r v i e w of the Technique 61 3.1.3 Advantages of G e l D o s i m e t r y 61 3.2 P o l y a c r y l a m i d e Gels 63 3.2.1 C o m p o s i t i o n 63 3.2.2 P o l y m e r i z a t i o n 64 3.2.3 Dose Response 69 3.3 P o l y m e r G e l D o s i m e t r y : A L i t e r a t u r e R e v i e w 69 3.3.1 F u n d a m e n t a l G e l Studies 70 3.3.2 G e l I m a g i n g M o d a l i t i e s 73 3.3.3 G e l A p p l i c a t i o n s 75 3.4 C T P o l y m e r G e l D o s i m e t r y 76 3.4.1 D e n s i t y a n d C T N u m b e r 77 3.4.2 L i t e r a t u r e R e v i e w 78 3.4.3 Advantages a n d L i m i t a t i o n s 80 3.5 Dose R e s o l u t i o n i n C T G e l D o s i m e t r y 80 3.5.1 A b s o l u t e Dose R e s o l u t i o n 81 3.5.2 R e l a t i v e Dose R e s o l u t i o n 82 C h a p t e r 4 M a t e r i a l s a n d M e t h o d s 83 4.1 G e l M a n u f a c t u r e 83 4.1.1 N i t r o g e n A t m o s p h e r e G l o v e B o x 84 4.1.2 O x y g e n M e t e r 84 4.1.3 G e l M a n u f a c t u r e Techn ique 86 4.2 G e l I r r a d i a t i o n 88 4.2.1 M e d i c a l L i n e a r A c c e l e r a t o r 88 4.2.2 I r r a d i a t i o n Techn ique 90 4.3 C T I m a g i n g 93 4.3.1 C T Scanner: G E H i S p e e d C T / i 93 4.3.2 C T I m a g i n g Techn ique 94 4.4 Image Process ing a n d D a t a A n a l y s i s 95 C h a p t e r 5 G e l C o m p o s i t i o n Studies 97 5.1 E x p e r i m e n t a l D e t a i l s 98 5.1.1 G e l C o m p o s i t i o n s 98 5.1.2 C T I m a g i n g 98 5.1.3 Image Process ing a n d D a t a A n a l y s i s 99 5.1.4 R a m a n Spectroscopy 100 5.2 Resu l t s a n d D i s c u s s i o n I: I n i t i a l Studies 100 5.2.1 C T P h a n t o m D e s i g n 100 5.2.2 Dose Response R e p r o d u c i b i l i t y 104 5.3 Resu l t s a n d D i s c u s s i o n I I : Effect of C r o s s - l i n k e r F r a c t i o n o n Dose Response 107 5.3.1 S e n s i t i v i t y 107 5.3.2 Dose R a n g e 110 5.4 Resu l t s a n d D i s c u s s i o n I I I : P A G D e n s i t y C h a n g e 112 5.4.1 T h e o r e t i c a l M o d e l 112 5.4.2 Effect of Cross - l inker F r a c t i o n 114 5.4.3 Effect of T o t a l M o n o m e r F r a c t i o n 119 5.5 C h a p t e r S u m m a r y 122 C h a p t e r 6 N o i s e a n d D o s e R e s o l u t i o n Studies 123 6.1 E x p e r i m e n t a l D e t a i l s 124 6.1.1 C T I m a g i n g 125 6.1.2 Image Process ing a n d D a t a A n a l y s i s 127 6.2 Resu l t s a n d D i s c u s s i o n 127 6.2.1 P h a n t o m D e s i g n 127 6.2.2 C T I m a g i n g Technique 129 6.2.3 V o x e l D i m e n s i o n s 132 6.2.4 P r e d i c t i n g Image No ise 135 6.2.5 Image U n i f o r m i t y 138 6.2.6 I m a g i n g T i m e 138 6.2.7 A c h i e v a b l e Dose R e s o l u t i o n 142 6.2.8 C o m p a r i s o n w i t h M R I a n d O C T G e l D o s i m e t r y 146 6.3 C h a p t e r S u m m a r y 147 C h a p t e r 7 D i g i t a l Image F i l t e r i n g Studies 149 7.1 E x p e r i m e n t a l D e t a i l s 150 7.1.1 B a c k g r o u n d G e l W o r k 150 7.1.2 Noise C h a r a c t e r i z a t i o n 151 7.1.3 D i g i t a l Image F i l t e r i n g 153 7.1.4 Image A n a l y s i s 154 7.2 Resu l t s a n d Dis cuss i on I: C h a r a c t e r i z a t i o n of Image No ise 156 7.2.1 C T P h a n t o m 156 7.2.2 G e l D o s i m e t e r 157 7.3 Resu l t s a n d Dis cuss i on I I : F i l t e r Per f o rmance 157 7.3.1 Dose R e s o l u t i o n 159 7.3.2 S p a t i a l Dose Integr i ty 163 7.4 C h a p t e r S u m m a r y 170 C h a p t e r 8 C o n c l u s i o n s 172 8.1 S u m m a r y of R e s u l t s 172 8.2 F u t u r e W o r k 177 A p p e n d i x A D i g i t a l Image F i l t e r s 178 B i b l i o g r a p h y 182 List of Tables 3.1 T h e s t a n d a r d P A G f o r m u l a t i o n 66 4.1 I m a g i n g parameters avai lable o n the G E H i S p e e d CT/i C T scanner. 95 5.1 P o l y m e r gel f ormulat ions used i n the P A G c o m p o s i t i o n a l studies . . . 98 5.2 I n t r a - b a t c h r e p r o d u c i b i l i t y of P A G dose response. A l l v ia l s were i r -r a d i a t e d to 8 G y . 105 6.1 S c a n parameters used to test the effects of scan p r o t o c o l o n image noise. P a r a m e t e r s i n b o l d face comprise the reference pro toco l . R e -cons t ruc t i on a l g o r i t h m names are s tandards for G E C T scanners. . . 126 6.2 A s u m m a r y of the measured q u a n t i t a t i v e effects of p h a n t o m size, C T scanning technique factors a n d voxe l size o n C T image noise 136 6.3 F r o m the results i n tab le 6.2, the percentage noise re la t ive to the ref-erence pro to co l (Re l . (JNCT) c a n ^ e der ived for a n y i m a g i n g pro toco l . G i v e n a noise measurement (Meas . VNCT) f ° r * n e reference pro to -co l , image noise can t h e n be ca l cu la ted for any C T i m a g i n g p r o t o c o l ( C a l c . <7/v c r). E x a m p l e s for low, average a n d h i g h noise protoco ls are p r o v i d e d 137 6.4 A compar i son of re lat ive dose resolut ions for a range of P A G dos ime-ter f ormulat ions used i n the i r l inear or "quas i - l inear " regions. A rep-resentative value of image noise, 0.3 H , was used i n a l l cases 143 6.5 A compar i son of re lat ive dose resolut ions for C T P A G d o s i m e t r y con-s t ra ined to meet a range of voxe l sizes. These results are not con-s t ra ined b y i m a g i n g t i m e . A 6 % T , 50 % C P A G was used 144 6.6 A compar i son of re lat ive dose resolut ions for C T P A G d o s i m e t r y con-s t ra ined to meet a range of voxel sizes for i m a g i n g a 1 L vo lume of gel (10 x 10 x 10 c m 3 ) w i t h i n a 1 hour t i m e l i m i t . A 6 % T , 50 % C P A G was used 145 6.7 A c o m p a r i s o n of the dose reso lut ion a n d t y p i c a l scan t imes for M R I , O C T , a n d x - r a y C T gel dos imetry . V o x e l sizes are g iven to fac i l i ta te compar i son . A l l quoted results are for P A G dosimeters 147 7.1 S u m m a r y of measured noise (<JNCT) a n d c a l c u l a t e d dose u n c e r t a i n t y (<TD) a n d dose reso lut i on w i t h 9 5 % confidence ( D ^ 5 % ) for t h e u n f i l -tered a n d a l l f i l tered images. ^Performance of the S U S A N fi lter is independent of m a s k size 162 7.2 S u m m a r y of H D R (at 750 c G y ) a n d r ight h a n d p e n u m b r a w i d t h (1000 to 500 c G y ) for the unf i l tered a n d a l l f i l tered images, tPer f o rmance of the S U S A N f i l ter is independent of m a s k size 168 7.3 S u m m a r y of ca l cu la ted dose area h i s t o g r a m values for the unf i l t ered a n d a l l f i l tered images. ^Performance of the S U S A N fi lter is i n d e p e n -dent of mask size 169 List of Figures 1.1 E x a m p l e s of percentage d e p t h dose curves ( P D D s ) i n water for a 10 M V p h o t o n b e a m ( spec t rum of energies, m a x i m u m 10 M e V ) (a), 9 M e V electrons (b) a n d 70 M e V protons (c) used i n r a d i a t i o n therapy. N o t i c e the h i g h l y p e n e t r a t i n g n a t u r e of the p h o t o n b e a m , the c o m -p a r a t i v e l y shal low d e p t h dose of the e lec tron b e a m a n d the d r a m a t i c dose fall-off at depth , t e r m e d the B r a g g peak, e x h i b i t e d b y the p r o t o n b e a m 5 1.2 A n overview of the process of m o d e r n l i n a c based r a d i a t i o n therapy. F o r c ompl i ca ted t reatment types or u n u s u a l c l i n i c a l cases, a t reatment p l a n ver i f i cat ion step is often requ ired . 3 D p o l y m e r gel d o s i m e t r y is a p r o m i s i n g t oo l w i t h the p o t e n t i a l to f i l l the current v o i d i n dosimeter technology avai lable to f i l l th i s r o l l 7 1.3 T h e basic pr inc ip le of c o n f o r m a l r a d i a t i o n therapy : f ie ld apertures are shaped to m a t c h the target i n the f ield 's b e a m eye v iew (a). F i e l d shap ing is achieved us ing b locks (b) or a m u l t i - l e a f c o l l i m a t o r ( M L C ) (c). C lear ly , c on formal r a d i a t i o n t h e r a p y provides greater n o r m a l tissue spar ing then i f l i m i t e d to j a w def ined, rec tangular fields (d). . 8 1.4 A n example of a m u l t i - l e a f c o l l i m a t o r ( M L C ) used i n con formal r a d i -a t i o n therapy. T h e p i c t u r e d M L C has 80, 1cm wide tungs ten leaves a n d is manufac tured by V a r i a n M e d i c a l Systems Inc. (Pa lo A l t o , C A ) . 9 1.5 A n example of a single in tens i ty m o d u l a t e d r a d i a t i o n t h e r a p y ( I M R T ) t reatment f ield. N o t e the h i g h l y m o d u l a t e d in tens i ty of r a d i a t i o n throughout the f ield. T h e scale is dose ( cGy) 10 1.6 A schematic d i a g r a m of the photoe lec tr i c effect. A n inc ident p h o t o n interacts w i t h a b o u n d a t o m i c e lectron p r o d u c i n g a fast e lectron a n d character is t i c r a d i a t i o n 13 1.7 A schematic d i a g r a m of C o m p t o n scat ter ing . A n inc ident p h o t o n interacts w i t h a "free" e lectron, transfers some energy to the e lectron a n d scatters w i t h the r e m a i n i n g energy. 14 1.8 A schematic d i a g r a m of p a i r p r o d u c t i o n . A n inc ident p h o t o n , hi/ > 1.022 M e V , interacts w i t h the e lectromagnet ic f ield of a n a t o m i c n u -cleus a n d is comple te ly absorbed . T h e result is p r o d u c t i o n of a n elec-t r o n a n d p o s i t r o n pa i r . T h e p o s i t r o n w i l l l a ter a n n i h i l a t e , p r o d u c i n g back to back 0.511 M e V photons 15 1.9 A schematic d i a g r a m of a F a r m e r type i o n chamber . Ions created w i t h i n the sensit ive vo lume are co l lected b y the electrodes. T h i s t y p e of i o n chamber is f requent ly used for c a l i b r a t i o n of r a d i a t i o n beams. 20 1.10 A schematic d i a g r a m of a d iode detector. It consists of two regions of doped s i l i con (n a n d p-type) operated i n a reverse bias . Inc ident r a d i a t i o n produces e lectron hole pairs a n d dose is recorded as a rise i n current t h r o u g h the detector 21 1.11 A schematic d i a g r a m of a thermoluminescent detector ( T L D ) . T L D dose measurement occurs i n two steps ( A a n d B ) , as i l l u s t r a t e d . I n step A , electrons a n d holes are t r a p p e d w i t h i n the T L D c r y s t a l after a n i o n i z i n g event. Step B is the emiss ion of l ight after r e c o m b i n a t i o n of the e lectron a n d hole w h e n the c r y s t a l is heated . T h e T L D l ight o u t p u t provides a measure of re lat ive dose 22 1.12 A p o l y a c r y l a m i d e gel ( P A G ) i r r a d i a t e d w i t h a stereotact ic r a d i o -surgery t rea tment . Some p o l y m e r gels, i n c l u d i n g th is one, undergo a v i s u a l change ( t u r n opaque) u p o n i r r a d i a t i o n 27 1.13 T h e dose d i s t r i b u t i o n for a stereotact ic rad iosurgery t reatment m e a -sured us ing C T p o l y m e r gel dos imetry . T h e d i s t r i b u t i o n is s h o w n i n the a x i a l (a), c orona l (b) a n d s a g i t t a l (c) planes. T h e l ines of equal dose (isodose lines) are for the ca l cu la ted dose d i s t r i b u t i o n . T h i s f igure is r eproduced f r o m [41] 28 2.1 Schemat i c i l lus t ra t i ons of the four classic generations of C T scanners. A first generat ion machine uses a single p e n c i l b e a m a n d a c o m b i n a -t i o n of t r a n s l a t i o n a l a n d r o t a t i o n a l m o t i o n (a). A second generat ion mach ine uses a narrow fan b e a m , m u l t i p l e detectors a n d a c o m b i n a -t i o n of t r a n s l a t i o n a l a n d r o t a t i o n a l m o t i o n (b). A t h i r d generat ion mach ine uses a large fan b e a m t h a t covers the entire object a n d the sole m o t i o n is the s imultaneous r o t a t i o n of the x - r a y t u b e a n d detec-tor a r ray (c). A f o u r t h generat ion m a c h i n e uses a so l id , s t a t i o n a r y r i n g of detectors a n d the on ly m o t i o n is r o t a t i o n of the x - r a y t u b e w i t h i n the r i n g (d) 35 2.2 T h e concept of p ro j e c t i on i m a g i n g , i l l u s t r a t e d for a g iven angu lar p o s i t i o n of source a n d detector 39 2.3 A n overview of a C T s cann ing sys tem 42 2.4 A schemat ic d i a g r a m of a d iagnost i c x - r a y t u b e 44 2.5 A schematic d i a g r a m of a so l id state detector 46 2.6 A n i l l u s t r a t i o n of the goal of r e c o n s t r u c t i o n i n x - r a y C T : to recon-s t ruct the d i s t r i b u t i o n of \x ( / ^ ' t h r o u g h ^4) i n a n object g iven m e a -sured p r o j e c t i o n samples (pi t h r o u g h ps) 50 2.7 A n example of i t erat ive image re cons t ruc t i on 51 2.8 A n i l l u s t r a t i o n of b a c k p r o j e c t i o n image recons t ruc t i on technique . T h e object , a u n i f o r m dot (a), is reconstructed b y b a c k p r o j e c t i n g the i n -tens i ty of measured pro ject ions across the ray l ines (b) 53 3.1 A s impl i f i ed overview of a s t a n d a r d gel d o s i m e t r y process. T h e gel is m a n u f a c t u r e d (a) a n d t h e n i r r a d i a t e d w i t h a t r ea tment for w h i c h a measured dose d i s t r i b u t i o n is des ired (b). T h e p o l y m e r i z e d gel is t h e n imaged to extract the dose i n f o r m a t i o n (c) a n d the images are processed to produce dose m a p s (d) 62 3.2 T h e chemica l s t ructure of a c r y l a m i d e (a), N, N'methylene b i s a c r y -l a m i d e (b) a n d their r a d i a t i o n p r o d u c t , c ross - l inked p o l y a c r y l a m i d e (c). N o t e the branched nature of the cross - l inked p o l y a c r y l a m i d e molecule 65 3.3 A sketch i l l u s t r a t i n g the nature of a s t a n d a r d P A G gel response to dose. T h i s classic response is e x p o n e n t i a l w i t h a quasi-linear reg ion at l ow doses often used for s impl i f i ed re lat ive dos imetry . It is i m p o r t a n t to note t h a t there are m a n y factors w h i c h affect th is dose response a n d w h a t is shown here is a sample response for i l l u s t r a t i v e purposes . 70 4.1 A schematic d i a g r a m showing the m a i n components of a d isso lved oxygen meter 85 4.2 S c i n t i l l a t i o n v ia ls a n d in-house m a d e p h a n t o m s used i n p r e p a r a t i o n , storage a n d i r r a d i a t i o n of p o l y m e r gels 87 4.3 A schematic d i a g r a m of a m e d i c a l l inear accelerator ( l inac ) . A s shown, the conf igurat ion i n the l i n a c head produces a p h o t o n b e a m . F o r e lectron b e a m p r o d u c t i o n the target is rep laced w i t h a s ca t ter ing f o i l 89 4.4 T h e set-up used for a l l p o l y m e r gel i r r a d i a t i o n s i n th is s tudy. E a c h gel v i a l was i r r a d i a t e d i n d i v i d u a l l y w i t h i n i ts c y l i n d r i c a l a c r y l i c p h a n t o m , p i c t u r e d i n figure 4.2, us ing the square i r r a d i a t i o n p h a n t o m as shown. 91 4.5 T h e t reatment p l a n designed for i r r a d i a t i o n of gel v ia l s . C T p l a n n i n g was used a n d the ca l cu la ted dose d i s t r i b u t i o n is s h o w n here super -i m p o s e d o n images of the v i a l i r r a d i a t i o n p h a n t o m i n three planes: a x i a l (a), c orona l (b) a n d s a g i t t a l (c). I n a l l v iews the 99 % isodose l ine (red) encompasses the gel v i a l vo lume 92 4.6 Schemat i c d i a g r a m of the H i L i g h t detector a r ray used i n the G E H i S p e e d CT/i C T scanner 94 5.1 T h e p h a n t o m used for C T i m a g i n g gel v ia l s . T h e p h a n t o m design al lows for s imultaneous i m a g i n g of a set of v ia ls a n d uses s tyro foam to improve image S N R 99 5.2 Effect of m a t e r i a l hous ing gel v ia l s o n the u n c e r t a i n t y i n measurement of v i a l N c r (ANCT)- T h i s u n c e r t a i n t y affects the prec i s i on of dose response curves ex t rac ted f r om these v i a l readings 101 5.3 R a w C T image of iner t , u n i r r a d i a t e d gel v ia l s o b t a i n e d u s i n g the s tyro foam C T i m a g i n g p h a n t o m s h o w n i n figure 5.1. N o t i c e the n o n -u n i f o r m i t y i n NCT w i t h i n the gel v ia l s 102 5.4 E x a m p l e of profiles t a k e n across a raw a n d b a c k g r o u n d s u b t r a c t e d C T image of gel v ia ls ob ta ined u s i n g the s tyro foam C T i m a g i n g p h a n t o m . T h e u n i f o r m i t y i n NCT for the b a c k g r o u n d s u b t r a c t e d image val idates the s tyro foam p h a n t o m technique 103 5.5 A n example of a f ina l averaged a n d b a c k g r o u n d s u b t r a c t e d image of i r r a d i a t e d gel v ia l s ob ta ined us ing the p h a n t o m p i c t u r e d i n f igure 5.1. N o t i c e tha t NQT is u n i f o r m w i t h i n the v i a l regions 103 5.6 C T dose responses measured for four independent batches of i d e n t i -ca l po lymer gel ( compos i t i on 6 % T , 50 % C ) . T h e r e p r o d u c i b i l i t y is excellent: the dose responses a l l agree w e l l w i t h i n the errors o n the fits. 106 5.7 T h e effect of % C o n the measured C T dose response of P A G gel dosimeters. A l l P A G s are 6 % T . E r r o r bars are der ived f r om a ± 0.2 H in te r -ba t ch r e p r o d u c i b i l i t y i n NQT measurements 108 5.8 T h e effect of P A G % C o n the c a l c u l a t e d sens i t i v i ty of P A G gel dos ime-ters. A l l P A G s are 6 % T 109 5.9 T h e extended C T dose response for a 100 % C , 3 % T P A G . B o t h a mono -exponent ia l ( x 2 = 0.010) a n d a l inear fit ( x 2 = 0.023) to 100 G y , are shown I l l 5.10 T h e densi ty change o c c u r r i n g i n P A G as a f u n c t i o n of dose for P A G s of v a r y i n g % C . T h e 0 a n d 100 % C P A G s are 3 % T a n d the r e m a i n i n g P A G s are 6 % T . E r r o r bars are der ived f r o m a i 0.2 H i n t e r - b a t c h r e p r o d u c i b i l i t y i n NQT measurements 115 5.11 T h e f rac t i on of monomer c o n s u m e d w i t h dose as de te rmined f r o m re lat ive R a m a n intensity . T h e 0 a n d 100 % C P A G s are 3 % T a n d the r e m a i n i n g P A G s are 6 % T . T h i s d a t a is ex t rac ted f r o m prev ious ly pub l i shed work a n d is shown here w i t h p e r m i s s i o n f r om the authors [86] 116 5.12 P o l y m e r y i e l d (weight f rac t i on of p o l y m e r f o rmed i n the gel) , as a f u n c t i o n of dose for P A G s of v a r y i n g % C . T h e 0 a n d 100 % C P A G s are 3 % T a n d the r e m a i n i n g P A G s are 6 % T 117 5.13 T h e effect of P A G % C o n the i n t r i n s i c dens i ty change, Appoiymer, o c c u r r i n g i n i r r a d i a t e d P A G 118 5.14 T h e effect of % T o n the dens i ty dose response of 0 % C P A G s . T h e sens i t iv i ty us ing 6 % T is twice t h a t for 3 % T , w i t h i n u n c e r t a i n t y i n the fits 120 5.15 T h e densi ty change per u n i t dose as a f u n c t i o n of % T for a 50 % C P A G . These results are e x t r a c t e d f r o m d a t a p u b l i s h e d prev ious ly a n d are shown w i t h permiss i on f r o m C . B a l d o c k [43] 121 6.1 T h e effect of p h a n t o m size o n C T image noise. T h e fit is exponent ia l . 128 6.2 Effects of scan technique (x - ray t u b e voltage, current a n d slice scan t ime) on C T image noise. N o t e t h a t different x -ax i s are p r o v i d e d for the three parameters 130 6.3 Image noise measured at a range of x - r a y t u b e loads achieved by changing each of t u b e voltage, current a n d slice scan t i m e i n d i v i d u a l l y f r o m the s t a n d a r d set of reference parameters (given i n tab le 6.1). E a c h parameter was v a r i e d over i t s f u l l opera t i on range as g iven i n tab le 6.1. T h e curve o b t a i n e d b y changing t u b e voltage is c lear ly m u c h steeper t h a n t h a t o b t a i n e d b y changing current or slice scan t i m e . T h i s indicates t h a t increas ing t u b e voltage is the most efficient means of l ower ing image noise t h r o u g h choice of scan pro toco l . . . . 131 6.4 T h e effect of choice of r e cons t ruc t i on a l g o r i t h m o n noise i n the resu l t -i n g C T image. T h e recons t ruc t i on a lgor i thms tested are s tandards for G E C T scanners 133 6.5 T h e effects of C T image voxe l d imensions on image noise. T h e i n -p lane d imens i on (p ixe l size) has a greater affect o n image noise t h a n does the d i m e n s i o n between i m a g i n g planes (slice th ickness a n d sep-arat ion) 134 6.6 S p a t i a l u n i f o r m i t y i n C T images as measured u s i n g 36 s p a t i a l l y d i s -t i n c t 25 x 25 p i x e l R O I s i n a single image (a) a n d 50 averaged images (b). B a c k g r o u n d s u b t r a c t i o n was app l i ed i n b o t h cases. E r r o r bars represent the s t a n d a r d d e v i a t i o n o n the m e a n for each R O I 139 6.7 T h e average t i m e required to o b t a i n a single slice image w h e n sequen-t i a l l y s cann ing m u l t i p l e slices u s i n g a single slice C T scanner. T h e t ime depends o n the l o a d p l a c e d o n the x - r a y t u b e a n d therefore o n C T scan technique . T h i s is due to requ i red t u b e coo l ing between slices. 141 7.1 C T images of the S R S gel a n d the c a l i b r a t i o n gel: a) a .s ingle , u n -processed S R S gel image, b) the processed (averaged a n d b a c k g r o u n d subt rac ted b u t unf i l tered) S R S gel image a n d c) the s i m i l a r l y p r o -cessed c a l i b r a t i o n gel image 151 7.2 Dose to NCT c a l i b r a t i o n curve o b t a i n e d f r o m the c a l i b r a t i o n gel shown i n figure 7.1c. NCT is the value above b a c k g r o u n d 152 7.3 C T image of the so l id water C T p h a n t o m used to character ize C T image noise. T h i s example inc ludes the a c r y l i c insert 153 7.4 P D F s for 6 u n i f o r m mater ia l s w i t h v a r y i n g C T contrast : a) p o l y p r o p y -lene, b) po lyethy lene , c) so l id water , d) a c r y l i c , e) tef lon a n d f) bone. A l l P D F s are fit w i t h Gauss ians . R e s u l t s ind i ca te that image noise is re lat ive ly independent of s igna l s t rength 155 7.5 P D F s for a b a c k g r o u n d region of a) the s ingle , unprocessed S R S gel image a n d b) the averaged a n d b a c k g r o u n d s u b t r a c t e d S R S gel image (shown i n figure 7.1a a n d b respect ive ly ) . B o t h P D F s are fit w i t h G a u s s i a n d i s t r i b u t i o n s . N o t e t h a t i n b) NCT is g iven re lat ive to the b a c k g r o u n d 156 7.6 R e s u l t of f i l t e r ing w i t h the m e a n fi lter ( M E A N ) u s i n g a 3 x 3, 5 x 5 a n d 7 x 7 p i x e l m a s k i n a) , b) a n d c) respect ively . 158 7.7 R e s u l t of f i l t e r ing w i t h the m e d i a n f i lter ( M E D I A N ) us ing a 3 x 3, 5 x 5 a n d 7 x 7 p i x e l m a s k i n a) , b) a n d c) respect ive ly 158 7.8 R e s u l t of f i l t e r ing w i t h the m i d p o i n t filter ( M I D P O I N T ) us ing a 3 x 3, 5 x 5 a n d 7 x 7 p i x e l m a s k i n a) , b) a n d c) respect ive ly 158 7.9 R e s u l t of f i l ter ing w i t h the m e a n adapt ive f i lter ( A D A P T I V E ) u s i n g a 3 x 3, 5 x 5 a n d 7 x 7 p i x e l m a s k i n a) , b) a n d c) respectively. . . 159 7.10 R e s u l t of f i l t er ing w i t h the a - t r i m m e d m e a n fi lter ( a - M E A N ) u s i n g a 3 x 3, 5 x 5 a n d 7 x 7 p i x e l m a s k i n a) , b) a n d c) respectively. . . 159 7.11 R e s u l t of f i l t er ing w i t h the s i g m a m e a n filter ( S I G M A ) u s i n g a 3 x 3, 5 x 5 a n d 7 x 7 p i x e l m a s k i n a) , b) a n d c) respectively. 160 7.12 R e s u l t of f i l ter ing us ing the smal lest un iva lue segment a s s i m i l a t i n g nucleus approach to noise r e d u c t i o n ( S U S A N ) . S U S A N filter perfor -mance is independent of mask size 160 7.13 F u l l H D R profiles a n d a m a g n i f i c a t i o n of the r ight h a n d p e n u m b r a (at r ight ) for the unf i l tered image a n d a l l filtered images for the 3 x 3 mask size 164 7.14 F u l l H D R profiles a n d a m a g n i f i c a t i o n of the r ight h a n d p e n u m b r a (at r ight ) for the unf i l tered image a n d a l l filtered images for the 5 x 5 mask size 165 7.15 F u l l H D R profiles a n d a m a g n i f i c a t i o n of the r ight h a n d p e n u m b r a (at r ight ) for the unf i l tered image a n d a l l filtered images for the 7 x 7 mask size 166 List of Abbreviations 3 D three d i m e n s i o n a l a - M E A N a l p h a - t r i m m e d m e a n f i lter A D A P T I V E adapt ive (Wiener ) f i l ter % C weight f rac t i on of m o n o m e r t h a t is crossl inker C T x - r a y c o m p u t e d t o m o g r a p h y Dj\ % re lat ive dose reso lut ion (%) H re lat ive dose reso lut ion w i t h 9 5 % confidence (%) Apgei change i n gel dens i ty w i t h i r r a d i a t i o n Appoiymer i n t r ins i c gel dens i ty change o c c u r r i n g per u n i t p o l y m e r f o rmed fm f rac t i on of m o n o m e r consumed F O V field of v i e w H D R h i g h dose region I M R T in tens i ty m o d u l a t e d r a d i a t i o n therapy L D R low dose region M E A N m e a n or "box" filter M E D I A N m e d i a n filter M I D P O I N T m i d p o i n t filter M L C m u l t i l e a f c o l l i m a t o r M R I magnet i c resonance i m a g i n g NCT C T n u m b e r i n H o u n s f i e l d u n i t s (H) (JNCT s t a n d a r d d e v i a t i o n of C T number ( image noise) ANCT change i n gel C T n u m b e r w i t h i r r a d i a t i o n O C T o p t i c a l c o m p u t e d t o m o g r a p h y % P p o l y m e r y i e l d P A G p o l y a c r y l a m i d e gel dos imeter P D D percentage d e p t h dose P D F p r o b a b i l i t y dens i ty f u n c t i o n S I G M A s i g m a fi lter S R S stereotact ic rad iosurgery S U S A N smallest un iva lue segmented a s s i m i l a t i n g nucleus ( S U S A N ) f i l ter % T weight f rac t i on of m o n o m e r i n a p o l y m e r gel T L D thermoluminescent detector Q D E q u a n t u m detector efficiency Acknowledgements I w o u l d l ike to t h a n k m y supervisor , D r . C h e r y l D u z e n l i for a l l o w i n g me w i t h freedom to explore m y ideas a n d p r o v i d i n g me w i t h expert ise a n d guidance w h e n I needed reminder of the larger p i c ture . I l ook f o r w a r d to our future co l laborat ions . T h e members of m y superv i sory commit tee , D r . J o h n A l d r i c h , D r . A l e x M a c K a y a n d D r . V e s n a Sossi , p r o v i d e d va luable comments t h r o u g h o u t th is w o r k a n d D r . Y v e s DeDeene e x a m i n e d th is thesis . T h e m e d i c a l i m a g i n g department a n d machine shop at V C C p r o v i d e d use of the C T scanner a n d a i d e d i n p h a n t o m c o n s t r u c t i o n a n d N S E R C a n d U B C p r o v i d e d fund ing . M y fami ly , friends a n d colleagues have p r o v i d e d suppor t , encouragement a n d he lp throughout th is w o r k a n d have m a d e th i s adventure enjoyable. M y warmest thanks to everyone for the t r i p s , the s k i i n g , the c l i m b i n g , the d inners a n d the coffee breaks. I n p a r t i c u l a r , t h a n k y o u to M o m a n d D a d w h o have always made m e believe I c ou ld a c c o m p l i s h a n y t h i n g . F i n a l l y , I want to t h a n k D r e w for b o t h his scientif ic advice a n d his cont inuous m o r a l suppor t . H e keeps me afloat. M I C H E L L E LOUISE HILTS The University of British Columbia, May 2005 Chapter 1 Introduction R a d i a t i o n is used to treat cancer b y de l iver ing a large l o ca l i zed dose of r a d i a t i o n to the disease vo lume , wh i l e spar ing s u r r o u n d i n g n o r m a l tissues. T h e process, t e r m e d r a d i a t i o n therapy, is a c o m p l e x m u l t i d i s c i p l i n a r y effort r e q u i r i n g accurate t u m o u r i m a g i n g , t reatment design, ver i f i cat ion a n d r a d i a t i o n del ivery. M o d e r n technology has developed r a p i d l y to meet the demands of r a d i a t i o n t h e r a p y a n d current t e ch -niques, such as in tens i ty m o d u l a t e d r a d i a t i o n t h e r a p y ( I M R T ) , can del iver h i g h l y complex three -d imens iona l (3D) dose d i s t r i b u t i o n s t h a t con form t i g h t l y to target vo lumes. However , the r a p i d development of r a d i a t i o n de l ivery techniques has not been m a t c h e d b y development of techniques for ve r i f y ing the accuracy of these t reatments . A s a p o t e n t i a l s o lu t i on to th is p r o b l e m , extensive research is c u r r e n t l y d i rec ted towards deve loping 3 D gel mater ia l s t h a t , w h e n i r r a d i a t e d , undergo c h e m -i c a l changes that c a n be measured u s i n g a n i m a g i n g moda l i ty . T h i s thesis involves invest igat ions i n the use of x - r a y c o m p u t e d t o m o g r a p h y ( C T ) for i m a g i n g a t y p e of these gels, p o l y a c r y l a m i d e gels ( P A G s ) . T h e effects of gel c o m p o s i t i o n o n the response of P A G to C T i m a g i n g are invest igated a n d a m o d e l is deve loped to fur -ther our u n d e r s t a n d i n g of the densi ty change t h a t occurs i n P A G u p o n i r r a d i a t i o n . C T i m a g i n g studies are per formed to develop o p t i m i z a t i o n strategies for i m a g i n g P A G a n d to determine achievable dos imeter character is t i cs . F i n a l l y , d i g i t a l image f i l t e r ing studies explore the idea of f i l t er ing C T images as a m e t h o d for i m p r o v i n g dose reso lut ion i n C T P A G dos imetry . T h r o u g h these bas ic studies , th i s w o r k has s igni f i cant ly furthered the f ield of gel dos imetry a n d paved the w a y for a c l i n i c a l l y v iab l e system i n the near future . 1.1 Radiation Therapy 1.1.1 O v e r v i e w C a n c e r remains one of C a n a d a ' s p r i m a r y l o n g - t e r m h e a l t h care crises. It is the second l ead ing cause of d e a t h a m o n g C a n a d i a n s a n d the l e a d i n g cause of d e a t h i n m i d d l e age (35 to 64 years) for b o t h sexes. E s t i m a t e s are 145,500 new cases a n d 68,300 deaths by cancer i n 2004, u p f r om previous years due , i n p a r t , to a n ag ing p o p u l a t i o n [1]. I n s u m m a r y , ~ 2 5 % of C a n a d i a n s w i l l die of cancer. T h e i m p a c t is also felt economica l ly : H e a l t h C a n a d a est imates a 14.2 b i l l i o n do l l a r per year cost to the heath care sys tem f r o m cancer re lated t reatments a n d h o s p i t a l i z a t i o n alone [1]-A p p r o x i m a t e l y 45 - 50 % of cancer pat ients receive some f o r m of r a d i a t i o n t h e r a p y t reatment for the i r disease. R a d i a t i o n t h e r a p y c a n be e i ther r a d i c a l , w i t h curat ive intent , or pa l l i a t i ve , w i t h the intent to manage p a i n a n d improve q u a l i t y of l i fe. T h e p r i m a r y goal of r a d i c a l r a d i a t i o n t h e r a p y is to del iver a t u m o u r i c i d a l dose of r a d i a t i o n to a w e l l def ined target vo lume (the disease) w h i l e spar ing s u r r o u n d i n g n o r m a l tissues. T h i s provides the greatest therapeut i c c o n t r o l w i t h a m i n i m u m level of m o r b i d i t y [2, 3]. T y p i c a l l y , r a d i a t i o n t h e r a p y is de l ivered b y d i r e c t i n g e x t e r n a l beams of r a d i -a t i o n in to a pat ient (external b e a m r a d i a t i o n therapy) or b y i n s e r t i n g or i m p l a n t i n g rad ioact ive sources into a pat ient (brachytherapy) . A n e x a m p l e of e x t e r n a l b e a m r a d i a t i o n therapy is the t reatment of prostate cancer u s i n g m u l t i p l e p h o t o n beams d i rec ted f r om var ious angles in to the pe lv is . A n example of b r a c h y t h e r a p y is the t reatment of c e rv i ca l cancer b y inser t ing a n a p p l i c a t o r in to the v a g i n a a n d v i a th is app l i ca tor p l a c i n g rad ioac t ive caes ium sources i n close p r o x i m i t y to the c e rv ix for a pre -determined l ength of t i m e . A l t h o u g h b r a c h y t h e r a p y is a n i m p o r t a n t component of r a d i a t i o n therapy cancer t rea tment , th is overview focuses on the more prevalent e x t e r n a l b e a m r a d i a t i o n therapy. E x t e r n a l b e a m r a d i a t i o n t h e r a p y t y p i c a l l y uses beams of photons or elec-trons , but protons a n d other heavy ions can also be used [4, 5]. T h e t y p e of r a -d i a t i o n selected for a t reatment depends large ly on the p a t t e r n of dose depos i t i on w i t h d e p t h a n d the scatter character ist ics of the r a d i a t i o n i n the m e d i u m [6]. T h e dose depos i t i on i n d e p t h a l ong the b e a m (section 1.2) is descr ibed b y the percentage d e p t h dose ( P D D ) curve a n d figure 1.1 shows example P D D s for photons , e lectrons a n d protons . N o t i c e the h i g h l y p e n e t r a t i n g na ture of the photons (a), the c o m p a r -a t ive ly shal low d e p t h dose of the electrons (b) a n d the very sharp dose fall -off at dep th , t e r m e d the B r a g g peak, e x h i b i t e d by the pro tons (c). These d i s t inc t P D D character ist ics f ind a p p l i c a t i o n i n various c l i n i c a l s i tuat ions . F o r example , due to the sharp dose fall-off, protons can be used for t r e a t i n g s m a l l eye t u m o u r s a d j a -cent to a h i g h l y c r i t i c a l s t ruc ture , the opt i c nerve. T o prov ide a su i tab le t r ea tment d e p t h , p r o t o n beams w i t h different energies are c o m b i n e d to spread out the n a r r o w h i g h dose region of the r a w B r a g g peak. I n contrast t o protons , h i g h energy p h o t o n beams can be used to t reat deep seated t u m o u r s i n the pe lv i s , such as c a r c i n o m a of the prostate . F u r t h e r e x p l a n a t i o n of these b e a m character is t i cs is p r o v i d e d i n sect ion 1.2 P r o t o n b e a m r a d i a t i o n t h e r a p y t y p i c a l l y requires a c y c l o t r o n for p r o t o n p r o -d u c t i o n a n d , as such, despite a t t rac t i ve d e p t h dose features (figure 1.1c), is less c o m m o n t h a n p h o t o n a n d e lectron r a d i a t i o n therapy. P h o t o n beams for r a d i a t i o n therapy can be p r o d u c e d f r o m a var ie ty of devices: o lder systems such as t h e r a p y x - r a y tubes (ki lovoltage rad ia t i on ) a n d cobalt -60 te le therapy u n i t s as w e l l as u l t r a -m o d e r n devices such as t o m o t h e r a p y u n i t s [7]. However , the most c o m m o n device for the p r o d u c t i o n of e x t e r n a l beams of b o t h p h o t o n a n d e lectron r a d i a t i o n is a m e d i c a l l inear accelerator , t e r m e d l inac . L i n a c s , descr ibed i n d e t a i l i n sect ion 4.2.1, are s m a l l l inear waveguide accelerators t h a t c a n produce r a d i a t i o n beams r a n g i n g i n energy f r om t y p i c a l l y 4 to 25 M e V . A n overview of the process of m o d e r n l i n a c based r a d i a t i o n t h e r a p y is g iven i n figure 1.2 [3]. A l l steps are c r i t i c a l for a n accurate t rea tment . A f t e r diagnosis a n d t reatment decisions are made , a pat ient w i l l undergo m e d i c a l i m a g i n g exams i n order to a l low accurate de l ineat ion of target vo lumes . T h i s w i l l c o m m o n l y involve a C T scan a n d i n cases where t u m o u r de l ineat i on is cha l leng ing , m a y also involve magnet i c resonance i m a g i n g ( M R I ) or nuc lear medic ine exams. W i t h a n accurate ly defined target vo lume, a t reatment p l a n is designed to del iver the prescr ibed dose of r a d i a t i o n to the target wh i l e m i n i m i z i n g n o r m a l t issue r a d i a t i o n dose. T h i s t rea t -ment p l a n n i n g is frequently per formed o n the pat ient C T images us ing , t yp i ca l l y , c o m m e r c i a l computer software t h a t provides tools for t reatment design a n d per -forms the dose ca l cu la t i on w i t h i n the pat ient anatomy. I n cases r e q u i r i n g c o m p l e x or u n u s u a l t reatment p lans , a n independent , measured ver i f i cat ion of the dose d i s -t r i b u t i o n p r o d u c e d d u r i n g the t reatment p l a n n i n g process m a y be requ i red before the f ina l stage, t reatment delivery. I n a l l cases, the b e a m qual i t ies of the l i n a c have been q u a l i t y contro l led t h r o u g h a series of da i ly , weekly, m o n t h l y a n d year ly b e a m checks a n d ca l ibrat ions . G e l dos imetry has the p o t e n t i a l to p l a y several i m p o r t a n t roles i n th i s p r o -cess. O n e role, t h a t takes place even before pat i ent t reatments can beg in , is i n the i m p l e m e n t a t i o n of new equipment or t r e a t m e n t techniques. T h i s process is ca l led commissioning a n d involves charac te r i za t i on a n d v a l i d a t i o n of equ ipment perfor -m a n c e a n d t reatment techniques t h r o u g h a de ta i l ed series of measurements . A n -other role is i n the ver i f i cat ion of specific pat ient t reatments t h a t are perhaps un ique or p a r t i c u l a r l y compl i ca ted . T h i s w o u l d require the development of gel p h a n t o m s w h i c h m i m i c pat ient a n a t o m y so that measured doses c o u l d be registered to c r i t i c a l Depth (cm) (a) Depth (cm) (b) 100 80 "2- 60 a 20 0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Depth (cm) (c) F i g u r e 1.1: E x a m p l e s of percentage d e p t h dose curves ( P D D s ) i n water for a 10 M V p h o t o n b e a m ( spec t rum of energies, m a x i m u m 10 M e V ) (a), 9 M e V electrons (b) a n d 70 M e V protons (c) used i n r a d i a t i o n therapy. N o t i c e the h i g h l y p e n e t r a t i n g na ture of the p h o t o n b e a m , the c o m p a r a t i v e l y shal low d e p t h dose of the e lectron b e a m a n d the d r a m a t i c dose fall -off at d e p t h , t e r m e d the B r a g g peak, e x h i b i t e d b y the p r o t o n beam. structures a n d target vo lumes w i t h i n the pat ient . I n b o t h these roles, alongside p r o v i d i n g 3 D dose measurements , another great advantage of gel d o s i m e t r y is t h a t a gel p h a n t o m c o u l d be r u n t h r o u g h the whole t reatment process, as o u t l i n e d above, jus t as i f i t was a pat ient . T h i s w o u l d result i n a h i g h l y va luable q u a l i t y assurance measure of the entire t r ea tment process. A s w i l l be discussed i n sect ion 1.3.3, i t is the increas ing c o m p l e x i t y of m o d e r n r a d i a t i o n t h e r a p y t reatments , as i n t r o d u c e d i n the fo l lowing sect ion, t h a t is t r u l y d e m a n d i n g the need for a dos imeter to f i l l a v o i d i n avai lable 3 D d o s i m e t r y techniques to a p p l y b o t h i n c o m m i s s i o n i n g a n d pat ient specific dose measurement . 1.1.2 M o d e r n T e c h n i q u e s P e r h a p s the greatest strides i n m o d e r n r a d i o t h e r a p y have been i n the a b i l i t y to con form dose d i s t r i b u t i o n s closely to target vo lumes . T h i s sect ion introduces the now re lat ive ly s t a n d a r d pract i ce of c o n f o r m a l r a d i a t i o n t h e r a p y a n d a more recent advancement t h a t is g a i n i n g widespread p o p u l a r i t y , in tens i ty m o d u l a t e d r a d i a t i o n therapy. C o n f o r m a l R a d i a t i o n T h e r a p y T r a d i t i o n a l r a d i a t i o n t h e r a p y consisted of square fields, shaped by l i n a c c o l l i m a t o r jaws (section 4.2.1). C o n f o r m a l r a d i a t i o n t h e r a p y differs i n t h a t a d d i t i o n a l b e a m modif iers are used i n order to con form r a d i a t i o n fields more closely to target vo lumes [8]. It developed fo l lowing the increased used of C T i n the p l a n n i n g process w h i c h a l lowed for bet ter target de l ineat ion . T h e basic p r i n c i p l e involves m a t c h i n g each b e a m aperture to the shape of the target vo lume w h e n pro jec ted in to the beam's eye v iew. T h i s b e a m shap ing is achieved u s i n g b locks or a m u l t i - l e a f c o l l i m a t o r ( M L C ) . F i g u r e 1.3 i l lustrates the pr inc ip l e of c o n f o r m a l r a d i a t i o n t h e r a p y u s i n g b o t h b locks a n d a n M L C a n d shows the increased s p a r i n g of n o r m a l t issue over use of t r a d i t i o n a l square t reatment fields. B l o c k s are c u s t o m m a n u f a c t u r e d for each Decision to pursue radiation therapy CT imaging for treatment planning Is the target ^clearly visible? No MRI and / or nuclear medicine imaging Register MRI or nuclear medicine images with CT images Yes Deliver treatment Yes Treatment follow-up ) F i g u r e 1.2: A n overview of the process of m o d e r n l i n a c based r a d i a t i o n therapy. F o r c o m p l i c a t e d t reatment types or u n u s u a l c l i n i c a l cases, a t r e a t m e n t p l a n ver i f i ca t i on step is often required . 3 D p o l y m e r gel dos imetry is a p r o m i s i n g t o o l w i t h the p o t e n t i a l to f i l l the current v o i d i n dos imeter technology avai lable to f i l l th i s r o l l . F i g u r e 1.3: T h e basic p r inc ip l e of con formal r a d i a t i o n therapy : f ield apertures are shaped to m a t c h the target i n the f ield 's b e a m eye v i ew (a). F i e l d shap ing is achieved u s i n g b locks (b) or a m u l t i - l e a f c o l l i m a t o r ( M L C ) (c). C l e a r l y , c on formal r a d i a t i o n t h e r a p y provides greater n o r m a l t issue spar ing t h e n i f l i m i t e d to j aw defined, rect -angu lar fields (d). F i g u r e 1.4: A n example of a m u l t i - l e a f c o l l i m a t o r ( M L C ) used i n conformal r a d i a t i o n therapy. T h e p i c t u r e d M L C has 80, 1 c m wide tungs ten leaves a n d is m a n u f a c t u r e d b y V a r i a n M e d i c a l Systems Inc. (Pa lo A l t o , C A ) . pat ient a n d each field a n d , b o l t e d to a tray , m a n u a l l y inserted into the head of the l inac for each t reatment . T h i s process is obv ious ly resource heavy a n d is t o d a y frequent ly replaced by use of a n M L C . A n M L C is a device m o u n t e d i n the l inac head t h a t consists of, t yp i ca l ly , 80 or 120 in terwoven tungs ten leaves that can move independent ly i n a n d out of the r a d i a t i o n b e a m . I n th is way, the M L C leaves can be pos i t i oned i n order to con form the r a d i a t i o n b e a m closely to the target , r educ ing n o r m a l tissue dose. A n 80 leaf M L C m a n u f a c t u r e d b y V a r i a n M e d i c a l Systems Inc. ( V a r i a n , P a l o A l t o , C A ) is p i c t u r e d i n f igure 1.4. I n t e n s i t y M o d u l a t e d R a d i a t i o n T h e r a p y P e r h a p s the greatest recent advancement i n t r ea tment de l ivery is the i n t r o d u c t i o n of in tens i ty m o d u l a t e d r a d i a t i o n therapy, or I M R T [9, 10]. A s e luded to i n the name , I M R T involves modulating r a d i a t i o n in tens i ty across b e a m apertures to produce beams of n o n - u n i f o r m fluence. T h i s feature makes I M R T un ique ly d i s t inc t f r om con formal r a d i a t i o n t h e r a p y (discussed above) i n w h i c h b e a m apertures are shaped , F i g u r e 1.5: A n example of a single in tens i ty m o d u l a t e d r a d i a t i o n therapy ( I M R T ) t reatment f ie ld. N o t e the h i g h l y m o d u l a t e d intens i ty of r a d i a t i o n throughout the f ield. T h e scale is dose ( c G y ) . but the r a d i a t i o n fluence within the aperture is u n i f o r m . T h e m o d u l a t e d r a d i a t i o n fluence is achieved b y s u m m i n g a series of sub-fields w i t h i n a single b e a m aperture . These sub-fields are t y p i c a l l y defined us ing a n M L C a n d can be del ivered either i n a step-and-shoot sequence (beam is off d u r i n g leaf mot ion ) or d y n a m i c a l l y (beam remains on throughout leaf m o t i o n ) . A m o d u l a t e d p a t t e r n of r a d i a t i o n intens i ty is o p t i m i z e d for each r a d i a t i o n f ield i n order to produce a dose d i s t r i b u t i o n con forming h i g h l y to the target vo lume u p o n s u m m a t i o n of a l l t reatment fields. A s a n example , the in tens i ty p a t t e r n for a single field of a c l i n i c a l I M R T treatment p l a n is shown i n figure 1.5. N o t i c e the h i g h degree of heterogeneity achieved throughout the f ield. S u c h fields a l l ow I M R T to produce complex dose d i s t r i b u t i o n s , p a r t i c u l a r l y valuable for c l i n i c a l sites where i r regu lar ly shaped t u m o u r vo lumes a n d c r i t i c a l s tructures are i n close p r o x i m i t y , such as head a n d neck cancers [11]. 1.2 Radiation Dosimetry Basics A comprehensive d iscuss ion of r a d i a t i o n d o s i m e t r y is w e l l b e y o n d the scope of th i s thesis. T h e fo l lowing sections i n t r o d u c e the q u a n t i t y absorbed dose (section 1.2.1) a n d the interact ions of r a d i a t i o n w i t h m a t t e r (section 1.2.2) w h i c h are centra l to r a d i a t i o n dos imetry . T h e reader is referred to several good texts for b o t h broader scope a n d greater d e p t h [6, 12, 13]. 1.2.1 A b s o r b e d D o s e T h e r a d i a t i o n dose absorbed b y a m e d i u m , referred to s i m p l y as dose (D), is defined b y D = D4± (1-1) dm where dEab is the m e a n energy absorbed into a m e d i u m of mass dm. T h e SI u n i t of dose is the G r a y ( G y ) , where 1 G y = 1 J k g - 1 . A n older u n i t of dose, the r a d , is sometimes s t i l l quoted : 1 r a d = 1 c G y . Ion i z ing r a d i a t i o n deposits energy i n m a t t e r t h r o u g h ion izat ions a n d e x c i t a -t ions of atoms a n d molecules . E l e c t r o n s are d i r e c t l y i o n i z i n g , but photons , b e i n g uncharged , transfer a f rac t i on of the i r energy to electrons i n the m e d i u m w h i c h t h e n deposit some of the i r energy as dose. P h o t o n s are therefore t e r m e d indirectly ioniz-ing a n d electrons directly ionizing r a d i a t i o n . T h e in termed ia te step i n the depos i t i on of dose f r om p h o t o n beams, the transfer of p h o t o n energy to set electrons i n m o t i o n i n the m e d i u m , is t e r m e d k e r m a (K). K e r m a , analogous to dose, is g iven b y K=d^L ( i .2 ) dm where d&tr is the m e a n energy transferred into a m e d i u m of mass dm. K e r m a also has u n i t s of G y . E l e c t r o n s of a g iven energy w i l l t rave l , depos i t ing dose i n a m e d i u m , for a n average distance t e r m e d the electron range before t h e y come to rest. A s a conse-quence, k e r m a a n d dose do not occur at the same l o c a t i o n i n a p h o t o n b e a m . Dose is absorbed s l i gh t ly d o w n s t r e a m of the i n i t i a l energy transfer or k e r m a . T h i s elec-t r o n range contr ibutes to the b u i l d u p of dose t h a t is observed at shal low depths for p h o t o n beams ( termed the " b u i l d - u p reg ion" ) , as seen i n figure 1.1a. S i m i l a r l y , the d e p t h of m a x i m u m dose i n a p h o t o n b e a m (referred to as dmax) is a p p r o x i m a t e l y equal to the range of the electrons set i n m o t i o n b y the photons . T h i s p a t t e r n of dose depos i t i on depends largely o n r a d i a t i o n energy a n d t y p e a n d , as descr ibed i n sect ion 1.1.1, is one of the most i m p o r t a n t character is t i cs i n d e t e r m i n i n g the c l i n i c a l s u i t a b i l i t y of a b e a m for r a d i a t i o n t h e r a p y app l i ca t i ons . 1.2.2 I n t e r a c t i o n s o f R a d i a t i o n w i t h M a t t e r T h i s sect ion discusses the basic interact ions w i t h m a t t e r of p h o t o n a n d e lectron beams used i n l i n a c r a d i a t i o n therapy. P h o t o n s T h e r e are four p h o t o n interact ions w i t h m a t t e r t h a t occur w i t h any s ignif icant p r o b a b i l i t y i n c l i n i c a l p h o t o n beams (~ 100 k e V - 25 M e V ) : coherent (Rayle igh) scat ter ing , photo -e lectr i c effect, C o m p t o n s ca t ter ing a n d pa i r p r o d u c t i o n . W i t h the except ion of coherent scatter ing , these interact ions result i n the p r o d u c t i o n of h i g h energy electrons a n d therefore, as discussed above, the depos i t i on of dose i n the m e d i u m . I n coherent scatter ing , a n inc ident p h o t o n sets a t o m i c electrons os c i l l a t ing , but the energy is r e - rad ia ted at the same frequency as the inc ident p h o t o n . Since there is no net energy transferred to the electrons, there is no dose depos i t i on . T h e basics of the three interact ions t h a t are responsible for dose depos i t i on are descr ibed here. T h e photoe lectr i c effect involves the i n t e r a c t i o n of a n inc ident p h o t o n w i t h a b o u n d a t o m i c e lectron. A s shown s chemat i ca l l y i n figure 1.6, the p h o t o n transfers a l l i ts energy to the e lectron, w h i c h is released w i t h h i g h k ine t i c energy f r o m its F i g u r e 1.6: A schematic d i a g r a m of the photoe lectr i c effect. A n inc ident p h o t o n interacts w i t h a b o u n d a t o m i c e lectron p r o d u c i n g a fast e lectron a n d character i s t i c r a d i a t i o n . shel l . T h e e lectron k ine t i c energy, KEe-, is g iven b y KEe- =hu-Eb (1.3) where hv is the energy of the inc ident p h o t o n a n d Eb is the e lectron b i n d i n g energy. T h e resu l t ing hole i n the inner shel l is f i l l ed b y a n outer she l l e lectron. T h i s results i n the p r o d u c t i o n of x - rays of energy hv = E b i - Ebo (1.4) where Eu a n d Ebo are the inner a n d outer shel l b i n d i n g energies respectively. These characteristic x-rays can further interact w i t h a t o m i c electrons, caus ing the i r emis-s ion as Auger electrons. T h e p r o b a b i l i t y of the photoe lectr i c effect o c c u r r i n g i n -creases w i t h m a t e r i a l a t o m i c n u m b e r (~ Z3) a n d decreases w i t h increas ing p h o t o n energy (~ 1/E3), w i t h d iscont inui t ies or absorption edges o c c u r r i n g at e lectron shel l b i n d i n g energies. I n C o m p t o n scat ter ing , as shown schemat i ca l ly i n f igure 1.7, a n inc ident p h o t o n interacts w i t h a n outer shel l a t o m i c e lectron w i t h b i n d i n g energy < < the energy of the inc ident p h o t o n such t h a t the e lectron can be considered "free". Some of the p h o t o n energy is t rans ferred to the e lectron a n d the p h o t o n scatters w i t h the r e m a i n i n g energy. T h e f r a c t i o n of energy t rans ferred to the reco i l e lec tron is fast electron incident photon / 9 \ s \ scattered photon F i g u r e 1.7: A schematic d i a g r a m of C o m p t o n s ca t ter ing . A n inc ident p h o t o n in ter -acts w i t h a "free" e lectron, transfers some energy to the e lectron a n d scatters w i t h the r e m a i n i n g energy. dependent o n the s cat ter ing angle a n d inc ident p h o t o n energy. T h e p r o b a b i l i t y of a p h o t o n C o m p t o n scat ter ing is g iven b y the K l e i n - N i s h i n a coefficient [12] a n d decreases s lowly w i t h inc ident p h o t o n energy (~ l/E) b u t is independent of m a t e r i a l a t o m i c number . P a i r p r o d u c t i o n (figure 1.8) involves the i n t e r a c t i o n o f a n inc ident p h o t o n w i t h the e lectromagnet ic field of a n a t o m i c nucleus . T h e p h o t o n is comple te ly absorbed a n d a n e lectron a n d p o s i t r o n p a i r are p r o d u c e d . A s a consequence, pa i r p r o d u c t i o n has a thresho ld p h o t o n energy of 1.022 M e V , twice the e lectron rest mass energy, 0.511 M e V . If the energy of the inc ident p h o t o n (hv) is > 1.022 M e V , the excess energy is shared as k ine t i c energy between the e lec tron (KE-) a n d p o s i t r o n T h e p o s i t r o n w i l l t r a v e l t h r o u g h the m e d i u m i o n i z i n g a n d e x c i t i n g a toms i n the m a t e r i a l . W h e n i t comes to rest, the p o s i t r o n w i l l a n n i h i l a t e w i t h a free e lectron e m i t t i n g two photons of energy 0.511 M e V , i n oppos i te d i rec t ions . T h e p r o b a b i l i t y of pa i r p r o d u c t i o n increases w i t h increas ing energy of the inc ident p h o t o n (above (KE+y. hv - 1 . 0 2 2 M e V = KE- + KE+. (1.5) / © electron at rest positron 0.511 MeV photon free electron O Q 0.511 MeV photon F i g u r e 1.8: A schematic d i a g r a m of p a i r p r o d u c t i o n . A n inc ident p h o t o n , hv > 1.022 M e V , interacts w i t h the e lectromagnet ic f ie ld of a n a t o m i c nucleus a n d is comple te ly absorbed . T h e result is p r o d u c t i o n of a n e lec tron a n d p o s i t r o n pa i r . T h e p o s i t r o n w i l l later a n n i h i l a t e , p r o d u c i n g back to back 0.511 M e V photons . the 1.022 M e V threshold) as ~ logE a n d also w i t h a t o m i c n u m b e r , as ~ Z2. T h e combined effects of these p h o t o n interact ions at tenuate the p h o t o n b e a m as i t passes t h r o u g h mat te r . P h o t o n b e a m a t t e n u a t i o n is g iven b y where I is the a t tenuated b e a m intens i ty , I0 is the inc ident b e a m intensi ty , / i is the t o t a l linear attenuation coefficient ( c m - 1 ) a n d x is p a t h l ength (cm). T h e t o t a l l inear a t t e n u a t i o n coefficient, JJL, is the s u m of the cross sections for pho to e lectr ic effect ( t ) , coherent a n d C o m p t o n s ca t te r ing (ocoh a n d <7i n c, respect ively) a n d p a i r p r o d u c t i o n ( « ) : P h o t o n b e a m a t t e n u a t i o n w i l l d e p e n d o n b e a m energy a n d also o n the n u m -ber of electrons per c m (and therefore densi ty ) of the m a t e r i a l . T o remove th is dens i ty dependence, p h o t o n b e a m a t t e n u a t i o n is t y p i c a l l y descr ibed b y the mass attenuation coefficient, fi/p ( c m 2 ; ? - 1 ) . A b s o r b e d dose (equat ion 1.1) is re lated to b e a m a t t e n u a t i o n by the mass energy absorption coefficient (fiab/p): I = i o e - i » (1.6) H = T + (Jcoh + Vine + K- (1.7) (1.8) b y D = * ( M ) (1.9) p where VP is the inc ident b e a m energy fluence, i n M e V c m 2 . E l e c t r o n s E l e c t r o n s at energies of interest i n r a d i a t i o n t h e r a p y (< 25 M e V ) interact w i t h m a t -ter p r i m a r i l y b y i o n i z a t i o n a n d e x c i t a t i o n a n d a r a d i a t i v e process ca l led b r e m s s t r a h l u n g . T h e interact ions are m e d i a t e d by the C o u l o m b forces between the charged par t i c l e a n d a tomic electrons ( ionizat ions a n d exc i tat ions ) a n d nuc le i (bremsstrahlung) . D u e to the i r s m a l l mass, electrons scatter easily, often w i t h l i t t l e loss of energy, a n d w i l l undergo m u l t i p l e scatters a n d changes of d i r e c t i o n as they pass t h r o u g h a m e d i u m . A s a side note, th is is w h y the B r a g g peak observed for heavier charged part i c les such as protons (figure 1.1c) is not observed for electrons (figure 1.1b). T h e energy lost b y a n e lectron b e a m as i t passes t h r o u g h m a t t e r is descr ibed b y the t o t a l e lectronic mass s t o p p i n g power, Stot- M a s s s t o p p i n g power is def ined as the k ine t i c energy lost per u n i t p a t h l ength , Stot=~ (1.10) p ax a n d has un i t s of M e V c m 2 g _ 1 . T h e t o t a l mass s t o p p i n g power is the s u m of s t o p p i n g powers account ing for the separate types of e lectron interact ions : mass i o n i z a t i o n a l (Sion) a n d mass rad ia t ive (Sraci) s t opp ing powers. Sion is a p p r o x i m a t e l y inverse ly p r o p o r t i o n a l to e lectron energy a n d decreases w i t h m a t e r i a l a t o m i c number . Srad increases d i rec t ly w i t h a t o m i c n u m b e r a n d increases more s lowly w i t h e lectron e n -ergy [12]. A s descr ibed above (section 1.2.1), absorbed dose is re lated to Sion, as th i s parameter describes the e lec tron energy t rans ferred to the m e d i u m t h r o u g h ion izat ions by: D = $Sion (1.11) where $ is the e lectron fluence, i n electrons c m - 2 . 1.3 Dosimeters in Radiation Therapy Devices that measure r a d i a t i o n dose, ca l l ed r a d i a t i o n dosimeters, are c r i t i c a l to p r o v i d i n g accurate r a d i a t i o n therapy. Dos imeters are used for tasks r a n g i n g f r o m c a l i b r a t i o n of l i n a c o u t p u t to m e a s u r i n g pat ient doses i n - v i v o . I n w h a t fol lows, sect ion 1.3.1, summar i zes some of the requ i red a n d desirable features of dosimeters for use i n r a d i a t i o n therapy. B r i e f descr ipt ions of the dosimeters most c o m m o n l y used i n r a d i a t i o n t h e r a p y are g iven i n sect ion 1.3.2 a n d sect ion 1.3.3 spec i f i ca l ly discusses 3 D dos imetry , p r o v i d i n g a br ie f i n t r o d u c t i o n to gel dos imetry . 1.3.1 D o s i m e t e r R e q u i r e m e n t s Specif ic dos imeter requirements w i l l depend o n the in tended a p p l i c a t i o n . However , there are dosimeter character ist i cs t h a t are un iversa l l y requ ired , or at least desirable , a n d these are s u m m a r i z e d here. S p a t i a l r eso lut i on is a n i m p o r t a n t dosimeter character is t i c . It determines the a b i l i t y of a detector to resolve the s p a t i a l d i s t r i b u t i o n of dose. T h e greater the s p a t i a l reso lut ion , the greater the a b i l i t y of the dosimeter to detect h i g h s p a t i a l v a r i a t i o n i n dose d i s t r i b u t i o n s . A s one m i g h t imagine , dosimeters w i t h h i g h s p a t i a l reso lut ion are increas ing ly necessary as r a d i a t i o n t h e r a p y t reatments become more complex a n d produce h i g h l y heterogeneous dose d i s t r i b u t i o n s . A g o o d example is I M R T , as descr ibed i n sect ion 1.1.2. F o r t r a d i t i o n a l r a d i a t i o n therapy, a 5 m m s p a t i a l reso lut ion is t y p i c a l l y sufficient, however m o d e r n techniques, such as I M R T , m a y require 3 m m or even 1 m m s p a t i a l reso lut ion [7, 14]. Dose reso lut ion , analogous to s p a t i a l reso lut ion , is a measure of the a b i l i t y of a dosimeter to resolve changes i n dose. Dose reso lut i on is def ined as the m i n i m u m detectable difference i n dose w i t h i n a c e r t a i n degree of confidence. Dos imeters w i t h h i g h dose reso lut i on are able to ac curate ly measure subt le changes i n dose. Dose reso lut ion is large ly de termined b y the sens i t iv i ty of the detector (i.e. s lope of i t s dose response) as w e l l as a d d i t i o n a l factors such as noise, w h i c h c ont r ibute to dose uncerta inty . T h e concept of dose reso lut i on w i l l be discussed i n greater d e t a i l as per ta ins to gel dos imetry , i n chapter 3. R e q u i r e d dos imeter dose reso lut ion is a p p l i c a t i o n specific. F o r example , ver i f i cat ion of a pat ient dose requires a dose reso lut i on of ~ 5 % [15], whereas a n absolute c a l i b r a t i o n of a l inac p h o t o n b e a m requires a dose reso lut ion of 2 % [7]. T h e r e are several a d d i t i o n a l i m p o r t a n t dosimeter character is t i cs . T h e a b i l i t y of a dosimeter to record the t o t a l a c c u m u l a t e d dose over a g iven i r r a d i a t i o n t i m e is c r i t i c a l for r a d i a t i o n therapy. T h i s a b i l i t y is t e r m e d dose integrating. Dos imeters used i n r a d i a t i o n therapy are designed w i t h a dose i n t e g r a t i n g mode i f possible . T h e result is t h a t a single dosimeter r e a d i n g is, i n some way, p r o p o r t i o n a l to the total dose del ivered to the device d u r i n g a n i r r a d i a t i o n . T i ssue equivalence is another va luable character is t i c . Since the dose depos i ted b y inc ident r a d i a t i o n is dependent o n the m a t e r i a l i r r a d i a t e d (refer to sect ion 1.2), dosimeters w h i c h are t issue l ike i n the i r response to r a d i a t i o n ( termed tissue equivalent) great ly s i m p l i f y dose measurement i n r a d i a t i o n therapy. W a t e r is general ly considered tissue equivalent for r a d i a t i o n dos imetry . W h e n measur ing dose us ing a dosimeter tha t is not water equivalent , such as a n i o n chamber , a series of correc t ion factors must be a p p l i e d to convert the measured dose to dose i n water [16-18]. O t h e r dosimeter character ist i cs are advantageous i n c e r t a i n s i tuat i ons . F o r example , the a b i l i t y to reuse a dosimeter is very p r a c t i c a l . T h i s is advantageous w h e n m a k i n g r a p i d consecutive measurements (for example w h e n a d j u s t i n g the out -p u t of a l i n a c beam) or w h e n one considers issues such as cost a n d waste. However there are s i tuat ions where a permanent dos imetr i c record m a y be desirable such as w h e n ver i fy ing the dose d i s t r i b u t i o n for a pat ient t reatment . A n o t h e r example is the a b i l i t y to measure a n a t o m i c a l l y relevant doses. Since i n c o n f o r m a l r a d i a t i o n t h e r a p y t reatment p lans are be ing t u n e d to pat ient specific anatomy, the a b i l i t y to measure the dose to that p a r t i c u l a r a n a t o m y is b e c o m i n g increas ing desirable. O f p a r t i c u l a r interest is the a b i l i t y to measure dose i n regions of h i g h u n c e r t a i n t y i n dose c a l c u l a t i o n such as a n a t o m i c a l inhomogeneit ies (e.g. lung) a n d i rregular surface topology. 1.3.2 C o m m o n D o s i m e t e r s U s e d i n R a d i a t i o n T h e r a p y T h e fo l lowing sect ion introduces the basic pr inc ip les a n d p r i m a r y app l i ca t i ons of the most c o m m o n types of dosimeters used i n r a d i a t i o n t h e r a p y hosp i ta ls for r a d i a t i o n b e a m a n d pat ient t reatment q u a l i t y assurance. T h e y are p o i n t dose dosimeters ( ion chambers , diodes a n d thermoluminescent dosimeters or T L D s ) , a n d a 2 D dosimeter , rad iograph i c f i l m . These summaries are large ly d r a w n f r o m several good books , [6, 7, 12] a n d the reader is referred to these references for more detai ls . Ion C h a m b e r s I on chambers come i n several forms for var ious app l i ca t i ons . O n e of the most c o m -m o n types for use i n r a d i a t i o n t h e r a p y is the t h i m b l e or F a r m e r - t y p e chamber as shown schemat i ca l ly i n figure 1.9. P l a c e d i n a b e a m of i o n i z i n g r a d i a t i o n , i o n i z a -t ions take place i n the a i r f i l led sensit ive vo lume of the chamber cav i ty a n d charge is co l lected o n the inner electrode. C h a r g e is measured b y connec t ing the chamber to a n electrometer. T h e chamber w a l l is made of " a i r equiva lent " p las t i c or graphi te (carbon) so as to not p e r t u r b the r a d i a t i o n b e a m . These i o n chambers are " open" so tha t t empera ture a n d pressure correct ions must be a p p l i e d i n order to k n o w the mass of a ir i n the chamber a n d enable c a l c u l a t i o n of absorbed dose de l ivered to the chamber . Ion chambers are the "go ld s t a n d a r d " i n c l i n i c a l r a d i a t i o n dos imetry a n d m a y be used to provide accurate , absolute , po in t dose measurements . O n e i m p o r t a n t use is i n the charac ter i za t i on of r a d i a t i o n t h e r a p y beams. F o r example , i o n chambers are used to ca l ibrate b e a m o u t p u t u s i n g w e l l def ined protoco ls f r o m the A m e r i c a n A s s o c i a t i o n of P h y s i c i s t s i n M e d i c i n e such as T a s k G r o u p ( T G ) 21, T G 5 1 or T G 2 5 [16-18]. I n a d d i t i o n , i o n chambers can be scanned i n d e p t h or across r a d i a t i o n outer electrode (inner wall) insulating material F i g u r e 1.9: A schematic d i a g r a m of a F a r m e r t y p e i o n chamber . Ions created w i t h i n the sensitive vo lume are co l lected by the electrodes. T h i s t y p e of i o n chamber is frequently used for c a l i b r a t i o n of r a d i a t i o n beams. beams us ing scanning water tanks i n order to measure percentage d e p t h dose curves (as shown previously , i n figure 1.1) a n d b e a m profiles. These b e a m character ist ics are requ ired by some c o m p u t e r i z e d t rea tment p l a n n i n g systems for dose c a l c u l a t i o n i n t reatment p l a n n i n g . A n o t h e r use of i o n i z a t i o n chambers is to provide a k n o w n , a c t u a l dose measurement at a po int to a i d i n ver i f i cat ion of complex t reatment p lans . F o r example , i n current pract i ce , a t y p i c a l I M R T p l a n ver i f i cat ion w o u l d inc lude one or more i on chamber po in t dose measurements a n d 2 D re lat ive f i l m measurements . However , recent work has i n d i c a t e d t h a t t r a d i t i o n a l F a r m e r - t y p e i o n chambers , due to their re lat ive ly large sensitive vo lumes (0.6 c m 3 ) , m a y not be able to measure b e a m character ist ics w i t h the h i g h s p a t i a l r eso lut i on required to accurate ly m o d e l I M R T treatments [19]. D i o d e s Diodes are semi -conductor detectors a n d those used for r a d i a t i o n dos imetry are general ly s i l i con p-n j u n c t i o n type diodes. A s shown schemat i ca l ly i n figure 1.10, they consist of s i l i con doped w i t h i m p u r i t i e s to prov ide pos i t ive a n d negat ive ly charged regions t e rmed p a n d n - t ype regions, respectively. T h e d iode funct ions i n reverse bias, w h i c h creates a dep le t i on layer at the j u n c t i o n of the p a n d n - t ype + V o F i g u r e 1.10: A schematic d i a g r a m of a diode detector. It consists of two regions of d o p e d s i l i con (n a n d p-type) operated i n a reverse bias . Inc ident r a d i a t i o n produces e lectron hole pairs a n d dose is recorded as a rise i n current t h r o u g h the detector. regions. W h e n i n a r a d i a t i o n b e a m , electrons a n d holes created t h r o u g h ion izat ions i n th is dep le t ion layer migrate to the p a n d n - t ype regions, respectively. T h i s current , measured w i t h a n e lectrometer , is p r o p o r t i o n a l to absorbed dose. D iodes , s imi lar to i o n chambers , are po in t dose or s cann ing I D dosimeters a n d are also often used to character ize t reatment beams. However , due to t empera ture a n d energy dependence of the response a n d a suscept ib i l i ty to r a d i a t i o n damage, they are not used for absolute d o s i m e t r y a n d b e a m c a l i b r a t i o n . T h e i r advantages inc lude a s m a l l size a n d h i g h sens i t i v i ty a n d they are f requent ly used to character ize c l i n i c a l e lectron beams. A s i o n chambers , diodes c a n be sealed water t ight a n d scanned i n a water t a n k to prov ide P D D a n d profi le measurements . I n a d d i t i o n , spec ia l ty sub -mi l l imeter d iode detectors ca l led M o s f e t s ™ ( M e d - T e c , Orange C i t y I A , U S A ) have been recent ly i n t r o d u c e d t h a t show promise for c l i n i c a l po in t dose measurements , p a r t i c u l a r l y i n - v i v o [20]. electron migration conduction band valence band O y hole migration F i g u r e 1.11: A schemat ic d i a g r a m of a thermoluminescent detector ( T L D ) . T L D dose measurement occurs i n two steps ( A a n d B ) , as i l l u s t r a t e d . I n step A , electrons a n d holes are t r a p p e d w i t h i n the T L D c r y s t a l after a n i o n i z i n g event. Step B is the emiss ion of l ight after r e c o m b i n a t i o n of the e lectron a n d hole w h e n the c r y s t a l is heated. T h e T L D l ight o u t p u t provides a measure of re lat ive dose. T h e r m o l u m i n e s c e n t D e t e c t o r s ( T L D s ) T h e r m o l u m i n e s c e n t detectors , ca l led T L D s , are so l id state crysta ls . T h e r e are m a n y types of T L D s , b u t a p o p u l a r one used i n r a d i a t i o n t h e r a p y is L i F doped w i t h M g or T i , ca l led T L D 1 0 0 . U s i n g the energy b a n d m o d e l , as shown schemat i ca l ly i n figure 1.11, T L D s have two energy bands separated b y a few e V : a valence energy b a n d , w h i c h confines r o o m temperature electrons, a n d a c o n d u c t i o n b a n d . T h e d o p i n g provides i m p u r i t i e s w h i c h l ead to l o c a l i z e d energy levels between these bands . U p o n i r r a d i a t i o n , e lectrons i n the valence b a n d ga in energy to a l low t r a n s i t i o n to the c o n d u c t i n g b a n d , step A i n figure 1.11. However , a f r a c t i o n of the electrons, p r o p o r t i o n a l to the dose de l ivered to the T L D , get t r a p p e d i n the i m p u r i t i e s . (Note that this process can occur w i t h holes as wel l ) . T h e t r a p p e d electrons can escape a n d r e t u r n to the c o n d u c t i o n b a n d b y h e a t i n g the T L D c r y s t a l , step B i n figure 1.11. These electrons recombine w i t h holes o n pass ing back to the valence b a n d , resu l t ing i n the emiss ion of l ight . T h e l ight emiss ion is measured , u s i n g a spec i f i ca l ly designed T L D reader, to prov ide a measure of r a d i a t i o n dose. T L D s are c o m m o n l y used i n r a d i a t i o n therapy for i n - v i v o pat ient dose m o n -i t o r i n g . T h e y are re lat ive dosimeters a n d must be c a l i b r a t e d to a k n o w n dose. I n fact, the response of i n d i v i d u a l T L D s are suff ic iently different t h a t each ch ip s h o u l d have i ts o w n c a l i b r a t i o n factor i n order to prov ide accurate dose measurements . However , th i s is general ly tedious i n pract i ce a n d t y p i c a l l y T L D s are c a l i b r a t e d i n batches of s i m i l a r response. T h i s process, i n a d d i t i o n to the careful h a n d i n g r e q u i r e d to achieve a reproduc ib le response, means that T L D s are on ly capable of p r o v i d i n g dose measurements w i t h i n ~ 5 % accuracy. However , i n other s i tuat i ons where i n -d i v i d u a l l y c a l i b r a t e d T L D s are used (such as for personnel r a d i a t i o n badges sent to n a t i o n a l laborator ies ) the accuracy is m u c h better ( t yp i ca l l y as good as 2 % ) . R a d i o g r a p h i c F i l m R a d i o g r a p h i c film is a re lat ive 2 D dosimeter . It consists of s i lver b r o m i d e crysta ls embedded i n ge la t in a n d spread o n a t ransparent p las t i c sheet. Inc ident r a d i a t i o n converts the s i lver b romide crysta ls to a t o m i c s i lver . T h e process of deve lop ing a n d fixing the film retains the s i lver i n exposed areas, b u t washes away the s i lver ions. T h e area of the f i l m where si lver remains appears dark , where no s i lver remains , clear. T h i s results i n a change i n the t r a n s m i t t a n c e of l ight t h r o u g h the film or film optical density. O p t i c a l density, measured w i t h a densi tometer , is p r o p o r t i o n a l to the r a d i a t i o n dose absorbed b y the film. T h e response is general ly non- l inear a n d film must be c a l i b r a t e d against k n o w n doses i n order to act as a dos imeter . F i l m response is also dependent o n b e a m energy, p a r t i c u l a r l y at low energies, due to presence of h i g h a t o m i c n u m b e r s i lver [21-23]. T h i s , i n c o m b i n a t i o n w i t h di f f icult ies i n c ontro l l ing f i l m process ing a n d poor i n t e r - b a t c h reproduc ib i l i t y , m e a n t h a t film accuracy is l i m i t e d to , at the best, ~ ± 3 %. Advantages however inc lude the a b i l i t y to measure r a d i a t i o n dose i n 2 D w i t h v e r y h i g h s p a t i a l reso lut ion a n d the permanence of the record. A s such, r a d i o g r a p h i c f i l m is used i n r a d i a t i o n t h e r a p y i n a l l s i tuat i ons r e q u i r i n g a 2 D dose measurement . S u c h s i tuat i ons arise b o t h i n t reatment u n i t q u a l i t y assurance a n d i n t reatment technique a n d p l a n ver i f i cat ion . F o r example , f i l m is c o m m o n l y used to measure the flatness of a r a d i a t i o n b e a m or the dose d i s t r i b u t i o n i n a plane t h r o u g h a c o m p l e x I M R T dose d i s t r i b u t i o n . 1.3.3 T h r e e D i m e n s i o n a l D o s i m e t r y M o t i v a t i o n Since r a d i a t i o n therapy treats t u m o u r vo lumes , dose d i s t r i b u t i o n s p r o d u c e d b y r a -d i a t i o n t h e r a p y have always been i n h e r e n t l y 3 D . However , the m o t i v a t i o n for 3 D d o s i m e t r y derives largely f r om i m p l e m e n t a t i o n of m o d e r n t reatment moda l i t i e s such as I M R T . F o r the homogeneous dose d i s t r i b u t i o n s de l ivered b y t r a d i t i o n a l r a d i a t i o n therapy, i t is possible to suff iciently character ize dose d i s t r i b u t i o n s b y measur ing the dose i n a p lane (2D) or even at a po in t . However , for m o d e r n con formal techniques, p o i n t a n d 2 D dos imetery is often inadequate . T h e r e are several reasons for th is di f -ference. A s descr ibed i n sect ion 1.1.2, techniques such as I M R T c a n produce h i g h l y heterogenous dose d i s t r ibut i ons . C l e a r l y , w h e n f u l l c h a r a c t e r i z a t i o n of such a 3 D dose d i s t r i b u t i o n is required , po int or even 2 D dose measurements are insuff ic ient. I n a d d i t i o n , the h i g h l y conformal nature of m o d e r n techniques i tse l f mot ivates 3 D dos imetry . F o r example , a dose d i s t r i b u t i o n t h a t is s cu lp ted to w r a p a r o u n d the s p i n a l cord , a s tucture h i g h l y sensitive to r a d i a t i o n , demands h i g h s p a t i a l reso lut ion d o s i m e t r y i n a l l three d imensions i n order to ensure no h i g h doses are de l ivered to the s p i n a l cord . T h i s is a l l the more c r i t i c a l w h e n one considers a new t r e n d towards dose escalation. Since n o r m a l t issue t o x i c i t y is often the l i m i t i n g factor i n prescr ib ing dose i n r a d i a t i o n therapy, the increased a b i l i t y t o con form t rea tment doses to the t u m o u r vo lume a n d spare s u r r o u n d i n g tissues means i t is possible to i n -crease or escalate the pres c r ip t i on dose w i t h o u t c reat ing u n d o n o r m a l tissue tox ic i ty . C l e a r l y , use of higher t reatment doses increases the r isks assoc iated w i t h inaccurate dos imetry a n d / or pat ient pos i t i on ing . F i n a l l y , a l t h o u g h p r o d u c i n g better dose d is -t r i b u t i o n s , the technology requ i red to del iver these dose d i s t r i b u t i o n s is increas ing ly complex . A s a result , i t is increas ing ly di f f icult for phys ic is ts to predic t t reatment doses a n d to have confidence i n the a b i l i t y of the technology to ac curate ly deliver the desired t reatment . F o r the i m p l e m e n t a t i o n a n d c o m m i s s i o n i n g of new techno l -ogy, a n accurate a n d re l iab le m e t h o d of ve r i f y ing dose d i s t r i b u t i o n s i n 3 D w o u l d be h i g h l y desirable. It is possible to a p p l y some of the s t a n d a r d dosimeters used i n r a d i a t i o n therapy (section 1.3.2) to 3 D dose measurement , b u t not w i t h o u t , i n some cases, severe l i m i t a t i o n s . T L D s , i o n chambers a n d diodes, w h i c h measure po int doses, have been used, p laced i n arrays , for 2 D p l a n a r d o s i m e t r y [24-26], a n d c o u l d po -t en t ia l l y be p laced i n a 3 D la t t i ce for p e r f o r m i n g 3 D dos imetry . However , i n order to achieve adequate s p a t i a l reso lut ion a large n u m b e r of dosimeters w o u l d be re -qu i red . F o r T L D s , th is w o u l d create a n e x t r e m e l y tedious a n d t i m e c o n s u m i n g task of keeping t rack of a n d r e a d i n g out a l l the i n d i v i d u a l T L D s . F o r i o n chamber a n d diodes, the prospect is even more unrea l i s t i c due to the radiosensi t ive e lectronics associated w i t h the dosimeters . I n a d d i t i o n , even i f c o n s t r u c t i o n of such systems were possible, the t issue equivalence o f the p h a n t o m w i l l be c o m p r o m i s e d by the presence of so m a n y dosimeters . F i l m , i n h e r e n t l y a 2 D too l , shows more promise since s tack ing sheets of film i n the 3rd d i m e n s i o n al lows for 3 D dose measurement [27, 28]. However , l ike T L D s , the s p a t i a l reso lut i on of th i s technique is l i m i t e d i n the s tack ing d imens i on due to b o t h tissue equivalence requirements a n d the i m p r a c t i c a l nature of processing m a n y films. I n a d d i t i o n , factors such as film process ing a n d ca l i b ra t i on , make accurate , reproduc ib le film d o s i m e t r y di f f icult to achieve. F i n a l l y , as hospi ta ls move increas ing ly to filmless, d i g i t a l futures , the cost of m a i n t a i n i n g a film processor p r i m a r i l y to p e r f o r m film d o s i m e t r y w i l l l i m i t the p r a c t i c a l i t y of 3 D film dosimetry . I n s u m m a r y , s t a n d a r d r a d i a t i o n t h e r a p y dosimeters are inadequate for 3 D dos imetry a n d a better so lu t i on is required . G e l D o s i m e t r y G e l dos imetry is cur rent ly the most p r o m i s i n g t o o l be ing developed for t r u l y 3 D dose measurement i n r a d i a t i o n therapy. T h e r e are m a n y different types of gel dosimeters , however a l l have the same u n d e r l y i n g p r i n c i p l e . R a d i o t h e r a p y gels consist of systems t h a t undergo measurable chemica l changes u p o n i r r a d i a t i o n t h a t are f ixed w i t h i n a gel m a t r i x . T h e two basic classes of gels c u r r e n t l y under inves t igat i on are Fr i cke a n d p o l y m e r gels. Fr i cke gels are based o n the Fr i cke chemica l dos imeter [29] i n w h i c h F e 2 + ions are converted to F e 3 + i n p r o p o r t i o n to absorbed dose [30]. P o l y m e r gels are based o n the p o l y m e r i z a t i o n of smal ler molecu lar const i tuents (monomers) u p o n i r r a d i a t i o n [31]. I n b o t h systems, the react ive chemicals are d ispersed i n a so l id gel m a t r i x ( typ i ca l ly ge lat in or agarose) so t h a t , at least for some given t i m e , reac t ion produc ts stay where they are created a n d the i r r a d i a t e d dose d i s t r i b u t i o n is s p a t i a l l y m a i n t a i n e d . T h i s thesis is concerned w i t h a p a r t i c u l a r type of p o l y m e r gel dosimeter , p o l y a c r y l a m i d e gels ( P A G s ) a n d a more deta i l ed d iscuss ion of p o l y m e r gel dos imetry is g iven i n chapter 3. F o r a l l gel dosimeters , recorded dose i n f o r m a t i o n is ex t rac ted f r o m gels t h r o u g h i m a g i n g . M a g n e t i c resonance i m a g i n g ( M R I ) was the first m o d a l i t y used for i m a g i n g b o t h classes of gels [30, 31] a n d is s t i l l used most frequently. M R I of i r r a d i a t e d Fr i cke a n d p o l y m e r gels show a measurable change i n l o n g i t u d i n a l a n d transverse r e l a x a t i o n rates ( i ? i a n d R2, respect ively) w i t h dose, so t h a t contrast i n M R I gel images is p r o p o r t i o n a l to dose. E x c e l l e n t dos imetr i c results have been ob ta ined us ing M R I gel dos imetry , however th i s requires a re la t ive ly h i g h degree of expert ise due to p o t e n t i a l artefacts t h a t resul t f r om magnet i c f ie ld inhomogeneit ies a n d h i g h sens i t iv i ty to gel t empera ture [32-35]. I n a d d i t i o n M R I is expensive a n d re la t ive ly inaccessible to most r a d i a t i o n t h e r a p y departments . A s a result , more re-cent ly a l ternat ive p o l y m e r gel i m a g i n g techniques have been proposed u s i n g o p t i c a l c o m p u t e d t o m o g r a p h y ( O C T ) [36, 37], R a m a n spectroscopy [38], x - r a y c o m p u t e d t o m o g r a p h y ( C T ) [39] a n d u l t r a s o u n d ( U S ) [40]. T h i s thesis is focused o n C T p o l y -F i g u r e 1.12: A p o l y a c r y l a m i d e gel ( P A G ) i r r a d i a t e d w i t h a stereotact ic radiosurgery t reatment . Some p o l y m e r gels, i n c l u d i n g th i s one, undergo a v i s u a l change ( t u r n opaque) u p o n i r r a d i a t i o n . mer gel dos imetry a n d a more deta i led discussion of C T is p rov ided i n chapter 2. Some of the e x c i t i n g features of gel dos imetry inc lude the inherent ly 3 D nature of the technique, the very h i g h spa t ia l reso lut ion of recorded dose i n f o r m a t i o n , the t issue equivalence a n d the a b i l i t y to construct a n t h r o p o m o r p h i c gel phantoms , even ones i n c o r p o r a t i n g inhomogeneit ies . F i g u r e 1.12 shows a n example of a p o l y m e r gel. T h i s p a r t i c u l a r gel is a P A G gel, i r r a d i a t e d w i t h a stereotact ic radiosurgery t reatment . T h e photo i l lustrates another pos i t ive feature of some p o l y m e r gels: a v i s u a l change u p o n i r r a d i a t i o n . T h e opaque regions are i r r a d i a t e d , the clear regions, u n i r r a d i a t e d . T h i s gel was imaged us ing C T a n d the result , shown i n figure 1.13 for three p e r p e n d i c u l a r planes, i l lustrates the p o t e n t i a l of gel dos imetry for 3 D dose ver i f i cat ion [41]. (a) F i g u r e 1.13: T h e dose d i s t r i b u t i o n for a stereotact ic radiosurgery treatment m e a -sured u s i n g C T p o l y m e r gel dos imetry . T h e d i s t r i b u t i o n is shown i n the a x i a l (a), coronal (b) a n d sag i t ta l (c) planes. T h e lines of equa l dose (isodose lines) are for the ca l cu la ted dose d i s t r i b u t i o n . T h i s figure is reproduced f rom [41]. 1.4 Thesis Goals T o s u m m a r i z e , p o l y m e r gel dos imetry is c u r r e n t l y be ing developed i n order to f i l l a 3 D dos imetry v o i d i n m o d e r n r a d i a t i o n t h e r a p y pract i ce . I n brief , p o l y m e r gel dosimeters are mater ia l s t h a t u p o n i r r a d i a t i o n undergo chemica l changes w h i c h record dose i n f o r m a t i o n i n 3 D . T h i s i n f o r m a t i o n is ex t rac ted b y i m a g i n g the dos ime-ter , a n d recently, x - r a y c o m p u t e d t o m o g r a p h y ( C T ) has been shown to be a p o t e n t i a l read-out m o d a l i t y due to a s m a l l dens i ty change t h a t occurs i n the dos imeter u p o n i r r a d i a t i o n [39, 42, 43]. T h e C T gel technique was f ound useful for l o c a l i z a t i o n of h i g h dose a n d h i g h dose gradient regions [41], b u t , i n general , th i s technique remains of l i m i t e d value for c l i n i c a l app l i ca t i ons due to , largely, poor dose reso lut ion . T h e poor dose reso lut i on is due to low sens i t i v i ty to dose (a weak C T number (NQT) to dose response) a n d noise i n the C T images. T h i s thesis studies the f u n d a m e n t a l n a t u r e of the response of p o l y a c r y l a m i d e gel ( P A G ) dens i ty to dose, techniques for C T i m a g i n g P A G a n d the p o t e n t i a l role of image process ing i n P A G dos imetry w i t h a n overa l l , u n i f y i n g goal of deve lop ing means of i m p r o v i n g the achievable dose reso lu t i on i n C T P A G dosimetry . A s s tated , th is goal is addressed v i a several avenues, a n d , as a result , is achieved t h r o u g h several independent studies. F i r s t , the u n d e r s t a n d i n g of the f u n d a m e n t a l na ture of the dens i ty change is furthered t h r o u g h s t u d i n g the effects of gel c o m p o s i t i o n o n the response of P A G to C T i m a g i n g . T h i s s t u d y develops a m o d e l to descr ibe the dens i ty change a n d characterizes the effect of gel c o m p o s i t i o n on i m p o r t a n t dos imetry parameters such as sens i t iv i ty , dose range a n d , u l t i m a t e l y , dose reso lut ion . T h e second set of studies are centered a r o u n d C T i m a g i n g a n d the development of i m a g i n g strategies for reduc ing noise a n d i m p r o v i n g dose reso lut i on w i t h i n the context of c l i n i c a l t i m e a n d voxe l size requirements . T h e t h i r d s t u d y investigates the propert ies of C T image noise a n d assesses the p o t e n t i a l of d i g i t a l image f i l t e r ing as a technique to fur ther improve dose reso lu t i on i n C T P A G dosimetry . B a c k g r o u n d i n f o r m a t i o n o n x - r a y c o m p u t e d t o m o g r a p h y a n d p o l y m e r gel d o s i m e t r y are p r o v i d e d i n chapters 2 a n d 3, respectively. Thes i s mater ia l s a n d methods are deta i led i n chapter 4. C h a p t e r 5 discusses the gel c o m p o s i t i o n studies a n d the m o d e l of p o l y m e r gel dens i ty change. T h e image noise a n d dose reso lut ion studies are deta i led i n chapter 6. T h e studies o n d i g i t a l image f i l t e r ing as a p p l i e d to gel dos imetry are discussed i n chapter 7. F i n a l l y , chapter 8 prov ides a s u m m a r y of results a n d discusses possible d irect ions for future work . Chapter 2 X-Ray Computed Tomography X - r a y c o m p u t e d t o m o g r a p h y i m a g i n g , c o m m o n l y k n o w n as C T , r evo lu t i on i zed d i a g -nost ic i m a g i n g w i t h i ts a b i l i t y t o produce cross-sect ional images of i n t e r n a l anatomy. C T is a n x - r a y i m a g i n g technique t h a t reconstructs images f r om m u l t i p l e pro jec -t ions of x - r a y t r a n s m i s s i o n d a t a co l lected a r o u n d the ob ject (or pat ient ) . T h i s thesis investigates the use of C T to extract dose i n f o r m a t i o n f r o m i r r a d i a t e d p o l y m e r gels a n d th is chapter provides detai ls of C T i m a g i n g w h i c h are pert inent to th is work . Research a n d development of C T techniques for use i n d iagnost i c i m a g i n g spans more t h a n 30 years. T h e reader is referred to m e d i c a l i m a g i n g physics t ex tbooks [44-46] or the recent, excellent C T t e x t b o o k b y H s i e h [47] for greater b r e a d t h a n d / o r dep th . Sec t i on 2.1 introduces the h is tory , t echno logy a n d f u n d a m e n t a l pr inc ip les of C T . M o d e r n C T scanners are descr ibed i n more d e t a i l i n sect ion 2.2. Sec t i on 2.3 p r o -vides a n i n t r o d u c t i o n to image recons t ruc t i on a n d deta i l s the re cons t ruc t i on m e t h o d used most f requent ly i n C T . F i n a l l y , sec t ion 2.4 discusses noise a n d artefacts i n C T i m a g i n g . 2.1 Introduction T h i s sect ion provides a n overview of the development of C T a n d introduces some f u n d a m e n t a l concepts. A br ie f h i s t o r y of C T is p r o v i d e d i n sect ion 2.1.1. Sect ion 2.1.2 introduces the different generations of C T scanner technology. T h e concept of p ro j e c t i on i m a g i n g , f u n d a m e n t a l t o C T , a n d the f o rm of C T scanner o u t p u t , C T numbers , are discussed i n sect ion 2.1.3. 2.1.1 A B r i e f H i s t o r y o f C T T h e invent i on of x - r a y c o m p u t e d t o m o g r a p h y represented a n enormous advance i n x - r a y i m a g i n g b y p r o v i d i n g sect ional images of i n t e r n a l anatomy. T h e signif icance of C T was recognized by the 1979 N o b e l P r i z e i n P h y s i o l o g y a n d M e d i c i n e be ing awarded to two pioneers of C T , A l l a n M . C o r m a c k a n d G o d f r e y N . Houns f i e ld . C o r m a c k der ived a m a t h e m a t i c a l t h e o r y of image re cons t ruc t i on i n the 1950s a n d tested th is theory o n w h a t is poss ib ly the first c o m p u t e d t o m o g r a p h y scanner ever b u i l t [48]. I n 1957, u s i n g a 6 0 C o source a n d a Geiger counter to per f o rm a l inear s canning exper iment C o r m a c k was able to ca l cu late the a t t e n u a t i o n coefficients of w o o d a n d a l u m i n u m i n a c y l i n d r i c a l p h a n t o m us ing his r e cons t ruc t i on theory. H e t h e n extended this work to a n a s y m m e t r i c a l p h a n t o m m a d e of a n a l u m i n u m r i n g (to represent the skul l ) s u r r o u n d i n g luc i t e (to represent soft tissue) w i t h i m b e d d e d a l u m i n u m plugs (to represent t u m o u r s ) . T h i s t i m e he per formed l inear scans at m u l t i p l e angular pro ject ions over 180 degrees. H e p u b l i s h e d the results to very l i t t l e a t t e n t i o n i n two papers i n 1963 a n d 1964 [47]. Interest ingly , i t was a desire to improve r a d i a t i o n therapy t reatments t h r o u g h m a p p i n g a t t e n u a t i o n coefficients w i t h i n the b o d y that o r i g i n a l l y m o t i v a t e d C o r m a c k ' s w o r k i n C T . Houns f i e ld is cred i ted w i t h the development of the first c l i n i c a l C T scanner w h i l e w o r k i n g for E l e c t r i c a n d M u s i c a l Industr ies ( E M I ) L t d . i n E n g l a n d 1 . T h e 1 Interestingly, it was high profits from the sale of Beatles records which allowed E M I to invest in new technologies (e.g. Hounsfield's C T scanner) in the 1960s. However, as a music company E M I had little experience in medical devices and were driven out of the C T market within a decade. f irst l a b o r a t o r y vers ion , b u i l d i n 1967, per f o rmed l inear scans o n a spec imen t h a t was ro ta ted i n 1 degree steps. D u e to a low in tens i ty g a m m a source a n d a h i g h l y inefficient m e t h o d of r e cons t ruc t i on (unl ike t h a t used by C o r m a c k ) , i t took 9 days to acquire the d a t a a n d reconstruct the first image [47]. Improvements to d a t a acqu i s i t i on a n d reconstruc t i on , use of a h i g h intens i ty x - r a y t u b e a n d s c i n t i l l a t i o n c r y s t a l detectors a l l c o m b i n e d to great ly improve the speed (4 1/2 minutes per image) a n d accuracy (0.5 %) of the sys tem [49]. I n 1971 the first c l i n i c a l C T scanner was ins ta l l ed at A t k i n s o n - M o r l e y H o s p i t a l i n the U K a n d the first pat ient image, a head scan per formed on O c t o b e r 4 t h 1971, showed a large cyst c lear ly v is ib le as a change i n image contrast f r o m the s u r r o u n d i n g t issue. A l t h o u g h C o r m a c k a n d Houns f i e ld are large ly hera lded as the pioneers of c l i n i c a l C T , work b y m a n y o ther researchers c o n t r i b u t e d to the i r achievements. A few key contr ibut ions are m e n t i o n e d here. J . R a d o n , a n A u s t r i a n m a t h e m a t i c i a n , proved i n 1917 t h a t a three -d imens iona l object c o u l d be reconstructed f r o m a n in f in i te set of i ts pro jec t ions [47]. R . N . B r a c e w e l l , a rad io -as tronomer , was the first to use this concept w h e n i n 1956 he cons t ruc ted a solar m a p f r o m pro j e c t i on data[44]. S i m i l a r to C o r m a c k , the pr inc ip les of C T were invest igated e x p e r i m e n t a l l y b y W . H . Oldendor f , a n a m e r i c a n neurologist , i n the 1950s [47]. W i t h the i n t r o d u c t i o n of the E M I scanner i n 1971, the C T r e v o l u t i o n i n d i -agnostic i m a g i n g h a d begun. T h e ensuing speed of i n t e g r a t i o n of C T into d iagnost i c medic ine is n o t h i n g short of a s t ound ing : b y 1977 at least 16 c o m m e r c i a l companies were m a r k e t i n g more t h a n 30 models of C T scanners a n d b y the new m i l l e n i u m more t h a n 5000 C T scanners were i n s t a l l e d i n the U n i t e d States alone [44]. C T scanner technology has evolved great ly i n th i s t i m e . T h e fo l l owing sect ion, 2.1.2, describes the classic "generat ions" of C T scanners. 2.1.2 G e n e r a t i o n s o f C T S c a n n e r s T h e evo lut i on of C T scanner technology is often descr ibed i n terms of generations. C T scanners are classif ied as f irst , second, t h i r d or f o u r t h generat ion based large ly o n the b e a m shape (penc i l or fan) a n d the s cann ing m o t i o n ( l inear a n d / o r r o t a t i o n a l ) . F i r s t to f o u r t h generat ion C T scanners, s h o w n s chemat i ca l l y i n F i g u r e 2.1, are descr ibed below. F i r s t G e n e r a t i o n T h e earliest C T scanners, such as t h a t m a n u f a c t u r e d by E M I i n the 1970s, are w h a t are now t e r m e d first generat ion C T scanners. These scanners use a single p e n c i l b e a m of x - rays coupled to a single detector. T h e x - r a y source a n d detector t rans la te together to collect pro ject ions l i n e a r l y across the object . T h e source a n d detector pa i r is t h e n r o t a t e d to another angle a r o u n d the object ( t y p i c a l l y by 1 degree) a n d the process repeated. F i g u r e 2.1a shows th is schemat ica l ly . T h e obvious disadvantage of th is technique is the l ong t i m e required for d a t a acqu i s i t i on , t y p i c a l l y 4 to 5 minutes , w h i c h frequent ly degrades images b y pat ient m o t i o n . S e c o n d G e n e r a t i o n Second generat ion C T scanners i m p r o v e d o n th i s ear ly design b y u s i n g a n a r r o w fan b e a m ins tead of the single c o l l i m a t e d p e n c i l b e a m . C o u p l e d to a n ar ray of detectors, th is technique, as s h o w n schemat i ca l ly i n F i g u r e 2.1b, a l lows for m u l t i p l e pro jec t ions to be co l lected at each source pos i t i on . T h e s cann ing m o t i o n remains s i m i l a r to the first generat ion systems: the source a n d detector a r ray are m o v e d l i n e a r l y t o collect p ro j e c t i on d a t a across the entire object a n d are t h e n r o t a t e d to another angu lar p o s i t i o n about the object a n d the l inear s cann ing process repeated. A n example is a n E M I system t h a t used a 30 degree fan b e a m a n d angular s cann ing increments of 10 degrees [47]. W i t h i m a g i n g t imes f r o m 20 to 60 seconds, these second generat ion systems s igni f i cant ly reduced the t i m e requ i red to generate a n image c o m p a r e d to (c) (d) F i g u r e 2.1: Schemat ic i l l u s t r a t i o n s of the four classic generations of C T scanners. A first generat ion machine uses a single p e n c i l b e a m a n d a c o m b i n a t i o n of t r a n s l a t i o n a l a n d r o t a t i o n a l m o t i o n (a). A second generat ion mach ine uses a narrow fan b e a m , m u l t i p l e detectors a n d a c o m b i n a t i o n of t r a n s l a t i o n a l a n d r o t a t i o n a l m o t i o n (b). A t h i r d generat ion machine uses a large fan b e a m t h a t covers the ent ire object a n d the sole m o t i o n is the s imultaneous r o t a t i o n of the x - r a y t u b e a n d detector array (c). A f our th generat ion mach ine uses a so l id , s t a t i o n a r y r i n g of detectors a n d the o n l y m o t i o n is r o t a t i o n of the x - r a y t u b e w i t h i n the r i n g (d). first generat ion scanners. T h i r d G e n e r a t i o n T h i r d generat ion scanners are s t i l l p o p u l a r today. L i k e the second generat ion m a -chines, these systems consist of a n x - r a y t u b e a n d a n oppos ing detector a r r a y t h a t rotate together about the object or pat ient . T h e difference is the detector arrays are larger concentr ic arrays that encompass the ent ire object w i t h i n the i r f ie ld of v iew. T h i s is shown schemat i ca l ly i n f igure 2.1c. T h i s design e l iminates the need for any t r a n s l a t i o n a l m o t i o n of the x - r a y tube a n d detector sys tem a n d great ly reduces i m a g i n g t i m e . I n c o r p o r a t i o n of m o d e r n s l ip r i n g technology for power a n d d a t a t ransmiss i on has produced t h i r d generat ion scanners w h i c h can image a single slice w i t h i n 0.5 s. These scanners are sometimes referred to as " ro ta te - ro ta te " machines since b o t h the x - r a y t u b e a n d the detector rotate . F o u r t h G e n e r a t i o n F o u r t h generat ion scanners e l iminate detector m o t i o n a l l together b y r ep lac ing the ar ray of detectors w i t h a 360° r i n g of detectors complete ly enclos ing the scanner bore. T h e x - r a y tube s t i l l produces a fan b e a m a n d rotates about the pat ient . A schematic d i a g r a m of th is geometry is s h o w n i n figure 2. I d . O n e advantage of th is type of sys tem is tha t i t a l lows for d y n a m i c c a l i b r a t i o n of detectors since each detector is exposed to a n u n a t t e n u a t e d x - r a y b e a m at some po in t d u r i n g a scan. However , since the detector r i n g c ircumference m u s t be qui te large, the requ i red number of detectors a n d associated electronics makes th is scanner des ign rather i m p r a c t i c a l . F o r example , a recently m a r k e t e d single slice f o u r t h generat ion mach ine contains 4800 detectors. P a r t i c u l a r l y i n l ight of the recent i n t r o d u c t i o n of m u l t i - s l i c e machines (as descr ibed i n sect ion 2.2.4) w h i c h w o u l d d e m a n d a great n u m b e r more detectors, these types of scanners are l ike ly to be phased out [47]. I n a d d i t i o n to these classic " four generat ions" of C T scanner technology, another un ique mach ine is w o r t h m e n t i o n i n g . A n u l t ra - fas t e lectron b e a m scanner was p r o d u c e d i n the 1980s i n efforts to image card iac m o t i o n . These scanners, sometimes t e rmed f i f th generat ion machines , use a sweeping h i g h speed e lectron b e a m t h a t , focused on a target r i n g , produces a n x - r a y fan b e a m . Since there is no mechan i ca l m o t i o n scan t i m e c a n be as fast as 50 ms [50]. 2.1.3 F u n d a m e n t a l P r i n c i p l e s o f C T C T i m a g i n g is a slice i m a g i n g technique . A series of 2 D a x i a l images of the object are p r o d u c e d a n d stacked together to create a 3 D image. E a c h 2 D image is created i n the same way. A s i l l u s t r a t e d b y the schematic d iagrams i n figure 2.1, x -rays are p r o d u c e d b y a n x - r a y tube a n d d i re c ted towards the object of interest . Detectors (or detector) o n the oppos i te side measure the x - rays t r a n s m i t t e d t h r o u g h the object . T h e measured x - r a y intensit ies are t e r m e d a projection. T h i s u n d e r l y i n g pr inc ip l e of projection imaging is the same as t h a t used i n convent iona l x - r a y i m a g i n g [51]. I n tha t case the detector is r a d i o g r a p h i c f i l m a n d a single p r o j e c t i o n produces a n image. B y contrast , i n C T i m a g i n g the x - r a y t r a n s m i s s i o n d a t a is co l lected for var ious angular pos i t ions about the object . I n m o d e r n C T scanners upwards of 1000 such pro ject ions are o b t a i n e d [45]. U s i n g speci f ical ly designed a lgor i thms , the pro ject ions are m a t h e m a t i c a l l y m a n i p u l a t e d i n order to reconstruct a n a x i a l slice of the object . T h e result is a n image where structures are not super imposed i n d e p t h as they are i n convent iona l x - r a y i m a g i n g . T h i s p r o v i d e d a n image c l a r i t y t h a t was a r evo lu t i on i n x - r a y i m a g i n g [52, 53]. T h e f ina l image o u t p u t is i n a f o rm ca l l ed C T numbers , w h i c h are r e la ted to object x - r a y a t t e n u a t i o n coefficients. T h i s sect ion detai ls the concept of p r o j e c t i o n i m a g i n g a n d defines C T numbers . Image reconstruc t i on f r om pro jec t ions is d iscussed i n a later sect ion, 2.3 P r o j e c t i o n I m a g i n g A s descr ibed i n chapter 1, x - rays are a t t enua ted as t h e y pass t h r o u g h a n object i n a n amount that depends o n the thickness a n d a t t e n u a t i o n coefficient of the m a t e r i a l traversed. I n a complex object , such as a pat ient , a single x - r a y b e a m w i l l traverse m a n y different tissues w i t h m a n y different a t t e n u a t i o n coefficients, \in. I n this case the t r a n s m i t t e d b e a m intensity , g iven i n equat i on 1.6 for a single becomes N I = I0exp (- A z ) (2.1) n=l where Arr is the thickness of each at tenuator . If A x approaches zero, equat i on 2.1 becomes I = IQexp(- fi(x)dx^) (2.2) where L is the l ength of the object traversed b y the x - r a y b e a m . A t each angular p o s i t i o n of the x - r a y t u b e a r o u n d the object , / , as g iven b y equat i on 2.2, is recorded a long the ray l ine j o i n i n g each detector to the x - r a y o u t p u t . T h i s concept is i l l u s t r a t e d s chemat i ca l ly i n figure 2.2. D i v i d i n g equat i on 2.2 b y the inc ident x - r a y b e a m intensity , I0, a n d t a k i n g the n a t u r a l l o g a r i t h m , a projection (p) is defined b y [47, 51] p=-ln{L)= J n{x)dx. (2.3) T h i s equat ion indicates tha t p ro j e c t i on d a t a encodes a l l the i n f o r m a t i o n necessary to produce a t t e n u a t i o n m a p images of the scanned object . T h e process by w h i c h th i s is done is ca l l ed image reconstruction. T h e r e are m a n y different methods of image reconstruc t i on a n d i t is a field of research i n i t s o w n r ight . T h e theory of reconstruc t i on techniques c o m m o n to x - r a y C T are discussed i n sect ion 2.3. C T N u m b e r s A s descr ibed above, C T scanners collect p ro j e c t i on d a t a of x - r a y t ransmiss i on t h r o u g h the scanned object . C o n t r a s t i n C T images is the result of differences i n the x - r a y projection F i g u r e 2.2: T h e concept of p r o j e c t i o n i m a g i n g , i l l u s t r a t e d for a g iven angular pos i -t i o n of source a n d detector. a t t e n u a t i o n coefficients of the var ious mater ia l s i n the object . Image p i x e l in tens i ty is a measure of the l inear a t t e n u a t i o n coefficient of the object , fi, re lat ive to the l inear a t t enuat i on coefficient of water , fiw. T h e in tens i ty is t y p i c a l l y expressed as a C T number (NCT ) i n H o u n s f i e l d u n i t s (H) a n d is g iven b y [47] N C T = 1000 x fLJ^E.. ( 2 .4) F o r example , NCT f ° r water , bone a n d l u n g are 0, ~ +1000 a n d ~ —1000 H , respectively. A C T image is a m a t r i x of C T numbers . F o r the purpose of v i e w i n g , each value of NCT is assigned a gray level . A v i e w i n g w i n d o w a n d level are defined i n order to o p t i m i z e the d i s p l a y contrast for the images. T h e w i n d o w defines the range of NCT tha t occupies the f u l l grayscale. T h e level defines the NCT at w h i c h the grayscale is centered. T h e w i n d o w a n d leve l c a n be ad justed i n order to h igh l i ght p a r t i c u l a r densities i n a n image , for example fine d e t a i l i n l u n g or bone [45]. 2.2 Modern C T Scanners T h i s sect ion describes m o d e r n C T scanners i n greater d e t a i l . T h e d iscuss ion fo-cuses on t h i r d generat ion machines . A sys tem overview is p r o v i d e d i n sect ion 2.2.1. Sec t ion 2.2.2 describes two m a j o r scanner components , the x - r a y tube a n d detector assemblies. Technique parameters avai lable for C T s cann ing are deta i led i n sect ion 2.2.3. Sect ion 2.2.4 h igh l ights two m o d e r n advances i n C T scanner technology, he-l i c a l a n d mul t i - s l i ce C T . T h e p a r t i c u l a r C T scanner used i n these studies , a G E H i S p e e d C T / i [54], is de ta i l ed i n the mater ia l s a n d methods , chapter 4. 2.2.1 S y s t e m O v e r v i e w T h e b lock d i a g r a m i n figure 2.3 out l ines the m a j o r components of a C T scanner a n d how they work together to produce C T images. T h e fo l l owing summar izes the steps requ i red to per form a C T scan. A scan t y p i c a l l y s tarts w i t h a scout image to define the region to be scanned. F o r th is process the x - r a y tube a n d detectors r e m a i n s t a t i o n a r y wh i l e the pat ient o n the couch moves t h r o u g h the scanner bore. T h i s produces a p l a n a r image s i m i l a r to a convent iona l x - r a y image that is used to define start a n d stop couch pos i t i on for the f u l l C T s can based on a n a t o m i c a l l a n d m a r k s . T h e couch is t h e n moved to the s tart p o s i t i o n , a n d the scanner is p r o g r a m m e d to per f o rm the scan. T h e r e m a y be pre-def ined scan protoco ls w h i c h v a r y w i t h the site be ing imaged . These protoco ls employ var ious scan parameters (e.g. x - r a y tube voltage, scan field of v i ew etc.) o p t i m i z e d for each a n a t o m i c a l site. Sec t ion 2.2.3 discusses these parameters i n d e t a i l . A f t e r p r o t o c o l select ion, the contro l c omputer sends c o m m a n d s to the sys tem (gantry, tab le , x - r a y tube , detectors a n d image generat ing software) to s tart the scan. T h e x - r a y g a n t r y ( i n c l u d i n g x - r a y t u b e a n d detector array) begins to rotate about the pat i ent , q u i c k l y reaching a constant opera t ing speed. T h e high-vol tage generator r a m p s u p to the desired voltage a n d m a i n t a i n s the prescr ibed voltage a n d current t h r o u g h o u t the scan. T h e x - r a y f lux traverses the pat ient , is detected b y x - r a y detectors a n d produces electronic s ignals . I n t a n d e m , the d a t a acqu is i t i on sys tem samples the detector o u t p u t at constant intervals a n d sends a d i g i t a l s ignal t o the image generat ion sys tem for processing. T h e image generat ion system reconstructs a n d post-processes the image w h i c h is sent to the operator computer for d isplay . 2.2.2 M a j o r C o m p o n e n t s A s descr ibed i n the overview, a C T scanner consists of m a n y components w h i c h must operate correc t ly to successfully produce a C T image. T h i s sect ion detai ls two of the most i m p o r t a n t components , the x - r a y t u b e a n d x - r a y detectors. X - R a y T u b e s X - r a y tubes generate the x-rays requ ired to p e r f o r m a C T scan a n d for m o d e r n h i g h speed s cann ing they must be able to do so cont inuously . F i g u r e 2.4 is a schemat ic high voltage generator x-ray tube gantry x-ray detectors couch data acquisition system control computer operator console computer + display image generating system F i g u r e 2.3: A n overv iew of a C T scanning system. d i a g r a m of a n i m a g i n g x - r a y t u b e as w o u l d be used i n C T . T h e m a i n components are a cathode a n d a n anode housed i n a glass or m e t a l envelope that sustains a v a c u u m env ironment . T h e cathode produces electrons a n d the anode is a target w h i c h u p o n b o m b a r d m e n t w i t h electrons produces a n x - r a y b e a m . B y a p p l y i n g a large current , a f i lament i n the cathode produces electrons b y t h e r m i o n i c emiss ion. A p o t e n t i a l difference between the cathode a n d anode accelerates the electrons t o -wards the anode. T h e speed of the electrons depends o n th is voltage, ca l led the tube voltage. T h e current p r o d u c e d i n the x - r a y tube , ca l l ed the tube current, is p r o p o r -t i o n a l to the number of electrons p r o d u c e d a n d depends o n the t e m p e r a t u r e of the cathode f i lament. T u b e current a n d t u b e voltage, as descr ibed i n sect ion 2.2.3, are two i m p o r t a n t technique parameters tha t can be v a r i e d to produce different scan protocols . A focusing cup at the cathode ensures the electrons h i t a w e l l def ined focal spot on the target . A s m a l l focal spot size is i m p o r t a n t for reduced geometr ic unsharpness i n C T images. X - r a y p r o d u c t i o n b y e lectron b o m b a r d m e n t , even w i t h use of a h i g h a t o m i c number target , is h i g h l y inefficient. A p p r o x i m a t e l y 9 9 % of the k ine t i c e lectron en -ergy is converted to t h e r m a l energy [47]. Several anode design features a c c o m m o d a t e th i s heat. T h e anode is t y p i c a l l y c ons t ruc ted of layers of m a t e r i a l to prov ide p h o t o n p r o d u c t i o n whi l e h a v i n g h i g h heat capacity . F o r example , some anodes use a m e t a l target (such as tungs ten - rhen ium) b r a z e d o n to graphi te for h i g h heat capac i ty [47]. I n a d d i t i o n , to increase the surface area b o m b a r d e d by the electrons wh i l e m a i n t a i n -i n g a s m a l l effective focal spot size, the anode is ang led a n d rotates [12]. A l t h o u g h most c l i n i c a l protoco ls can be per formed w i t h o u t t u b e overheat ing , C T scanners t y p i c a l l y employ a lgor i thms that force either technique r e d u c t i o n (i.e. reduced t u b e voltage a n d / or current) or a coo l ing p e r i o d i f a n i m a g i n g p r o t o c o l w i l l result i n heat t h a t m a y damage the x - r a y tube . F i g u r e 2.4: A schematic d i a g r a m of a d iagnost i c x - r a y tube . X - R a y D e t e c t o r s X - r a y detectors i n C T scanners are e i ther gas-f i l led i o n chambers or so l id-state detectors. T h e y are selected to have h i g h efficiency, short response t i m e a n d a stable response. T h e inert gas x e n o n is frequently used i n gas-f i l led i o n chambers [44]. U p o n i r r a d i a t i o n xenon ionizes to produce pos i t ive xenon ions a n d electrons. T h e electrons are co l lected at plates w i t h h i g h pos i t ive voltage (~ 500 V ) p r o d u c i n g a current . T h e voltage is h i g h enough t h a t the detector response is r a p i d a n d there is l i t t l e i o n r e c o m b i n a t i o n but not so h i g h t h a t a non- l inear avalanche type response w o u l d occur [47]. I n such a sys tem the current is l inear ly p r o p o r t i o n a l to the t o t a l energy of the absorbed x - rays . A m a j o r d isadvantage of x e n o n i o n chamber detectors is the i r l ow q u a n t u m detector efficiency ( Q D E ) , the percentage of inc ident photons a t tenuated by the detector . T h i s is due to the low densi ty of xenon gas w h i c h al lows m a n y photons to pass t h r o u g h undetected . E v e n for h i g h pressure gas systems, Q D E is on ly 60 to 70 %. However , the m a j o r advantage is l ow cost a n d lower end m o d e r n C T scanners s t i l l use th is t y p e of detector . I n most m o d e r n C T scanners so l id-state s c i n t i l l a t i o n crystals a n d p h o t o d i -odes have replaced xenon i o n chambers as detector systems. A schematic d i a g r a m of a so l id state detector is shown i n figure 2.5. S m a l l b locks of a s c i n t i l l a t i n g m a -t e r i a l separated b y reflective m a t e r i a l are coupled to photod iodes . M a n y different s c i n t i l l a t i o n mater ia l s have been used, for example N a l , B G O a n d C d W 0 4 . E n e r -getic electrons p r o d u c e d w h e n i n c o m i n g x - rays ion ize the s c i n t i l l a t i n g m a t e r i a l exc i te other electrons w i t h i n the m a t e r i a l . These electrons, u p o n r e t u r n i n g to the ir g r o u n d state , emit l ight . T h e l ight o u t p u t is recorded b y the photod iodes as electric s igna l . T h e reflective m a t e r i a l serves to d irect the e m i t t e d l i ght towards the photodiodes . Q D E for these so l id-state systems is m u c h i m p r o v e d over the x e n o n i o n chambers . F o r example , the HiLight® sol id-state detector m a r k e t e d b y G E boasts a Q D E of 98 to 99.5 % [47]. However , the overa l l detector efficiency (i.e. % inc ident photons a c t u a l l y collected) is s igni f i cant ly lower due to a geometr ic efficiency of ~ 80 %. O t h e r per formance parameters to consider w i t h so l id -state detectors inc lude after-glow, suscept ib i l i ty to r a d i a t i o n damage , t h e r m a l s t a b i l i t y a n d response un i f o rmi ty . 2.2.3 T e c h n i q u e P a r a m e t e r s T h e r e are m a n y technique parameters w h i c h are selected w h e n p e r f o r m i n g a C T scan to prov ide the best images for each a p p l i c a t i o n . P a r a m e t e r choice affects the resu l t ing image noise, contrast , sharpness, s p a t i a l r e so lu t i on a n d field of v iew. T h e y also affect pat ient dose, a n i m p o r t a n t cons iderat i on i n C T i m a g i n g [55, 56]. T h i s sect ion defines these parameters a n d out l ines t h e i r effects o n images a n d pat ient dose. T h e o p t i m i z a t i o n of C T technique parameters for gel dos imetry is discussed i n chapter 6. scintillating material material photo diodes F i g u r e 2.5: A schemat ic d i a g r a m of a so l id state detector. T u b e V o l t a g e T u b e voltage, often s i m p l y t e r m e d kV, is the vo l tage app l i ed between the cathode a n d anode i n a n x - r a y tube . A s descr ibed i n sect ion 2.2.2, t u b e voltage determines the energy of electrons i m p a c t i n g the target . Increasing t u b e voltage results i n increased average p h o t o n energy a n d , since e lectron rad ia t ive s t opp ing power i n -creases w i t h e lectron energy for target mater ia l s such as tungs ten [12], p h o t o n y i e l d per e lectron also increases. A s discussed i n d e t a i l i n sect ion 2.4.1, s t a t i s t i ca l noise i n C T images decreases w i t h increas ing n u m b e r of photons p r o d u c e d (JV). A s a result , increas ing t u b e voltage decreases image noise. However , the increase i n av-erage p h o t o n energy w i t h increased t u b e voltage c a n reduce the contrast between structures w i t h different a t o m i c numbers (e.g. bone a n d soft t issue) . T h i s is due to the reduced p r o b a b i l i t y of the photoe lec t r i c effect w i t h increas ing b e a m energy, as descr ibed i n chapter 1. Increas ing t u b e voltage also increases pat ient dose since dose increases w i t h N [53, 57]. F o r example , g iven a l l other scan parameters the same, a five fo ld increase i n pat ient dose was measured between scans us ing 80 a n d 140 kV t u b e voltage sett ings [56]. T u b e C u r r e n t T u b e current , often s i m p l y t e r m e d mA, is the current of electrons acce lerat ing be-tween the cathode a n d anode i n a n x - r a y tube . A s descr ibed i n sect ion 2.2.2, t u b e current is p r o p o r t i o n a l t o the n u m b e r of electrons h i t t i n g the target per second. T u b e current does not affect the e lec tron energy a n d therefore the n u m b e r of pho -tons p r o d u c e d (N) is p r o p o r t i o n a l t o t u b e current . Hence , as above, increas ing tube current decreases image noise a n d increases pat ient dose [47, 57]. A t y p i c a l t u b e current is 200 m A . However , w h a t is c o m m o n l y quoted is m A s , the p roduc t of t u b e current a n d slice scan t i m e . Slice S c a n T i m e Slice scan t i m e is the t i m e i t takes for the C T g a n t r y to rotate about the pat ient a n d the t i m e required to scan a single image slice. Increas ing slice scan t i m e increases N, decreases image noise a n d increases pat ient dose. Increas ing slice scan t i m e also increases the t i m e requ i red to complete a C T scan. A t y p i c a l slice scan t i m e is 1 second a n d i t is c o m m o n l y quoted c o m b i n e d w i t h tube current , for e.g. 200 m A s . F i e l d o f V i e w F i e l d of v iew ( F O V ) determines the p h y s i c a l area i m a g e d b y the C T scanner. F o r example , for a head scan m a y used a s m a l l F O V c a n be used w h i l e for a n a b d o m i n a l scan a large F O V is required . T h e advantage of r e d u c i n g F O V is a decreased p i x e l size a n d therefore i m p r o v e d i n - p l a n e (x,y) s p a t i a l reso lut ion . F o r example a t y p i c a l head scan m a y use a F O V of 25 x 25 c m 2 , w h i c h provides a p i x e l d i m e n s i o n of < 0.5 m m for a 512 x 512 p i x e l image. However , th i s decrease i n p i x e l size also impl i e s a decrease i n the number of photons detected per p i x e l , r e su l t ing i n a n increase i n s ta t i s t i ca l image noise. Slice Index Slice index is the couch increment made between successive C T slices. I n o therwords i t is the distance between a x i a l images. A s such , increas ing slice index results i n decreased s p a t i a l reso lut ion i n the out -o f - image p lane d i rec t i on . T h i s l o n g i t u d i n a l d i re c t i on w i l l be referred to f r om th is po in t o n as the z d i re c t i on . A t y p i c a l slice index is 3 or 5 m m . Slice T h i c k n e s s Slice thickness , de te rmined b y b e a m c o l l i m a t i o n , is the slice d imens i on i n the z d i rec t i on . I n f o r m a t i o n w i t h i n the slice thickness is compressed in to a single 2 D image. A l o n g w i t h p i x e l d imens ion , slice thickness defines the size of image voxels. Increasing slice thickness increases voxe l size w h i c h increases the number of p h o t o n interact ions per voxel a n d decreases image noise. However , increas ing slice th ickness can reduce image q u a l i t y by i n t r o d u c i n g vo lume averaging artefacts , as descr ibed i n sect ion 2.4.2. I n a d d i t i o n , increased slice thickness can reduce s p a t i a l reso lut i on since slice index is t y p i c a l l y m a t c h e d to slice thickness i n C T scanning . A t y p i c a l slice thickness is 3 or 5 m m . R e c o n s t r u c t i o n A l g o r i t h m C T scanners have several b u i l t i n re cons t ruc t i on a l g o r i t h m s to choose f rom. These a lgor i thms t y p i c a l l y differ i n the fi lters they a p p l y to the pro j e c t i on d a t a w h i c h serve to , for example , enhance fine d e t a i l or improve contrast reso lut ion i n the f ina l image. T h e theory of f i l tered image re cons t ruc t i on is discussed i n more d e t a i l i n sect ion 2.3. These a lgor i thms can have d r a m a t i c a l l y different effects o n image noise a n d image sharpness a n d are selected to benefit a p a r t i c u l a r a p p l i c a t i o n . 2.2.4 R e c e n t A d v a n c e s T h e d iscuss ion presented thus far has focused o n single slice, a x i a l C T i m a g i n g . T h i s was done b o t h for c l a r i t y a n d to d e t a i l the t y p e of s canning used i n the e x p e r i m e n t a l w o r k i n th is thesis. T h i s sect ion describes two recent advances to C T technology w h i c h have been w i d e l y adopted , he l i ca l a n d mul t i - s l i c e C T . I n he l i ca l C T the couch moves t h r o u g h the scanner bore at constant speed w h i l e the gantry is r o t a t i n g a n d the pro j e c t i on d a t a is be ing co l lected. I n th i s way a he l i x of d a t a covers the scanned vo lume . T h i s is i n contrast to the set of discrete a x i a l slices ob ta ined w i t h convent ional s tep-and-shoot C T scanning . T h e r a t i o of the tab le movement per slice to the slice thickness is ca l l ed the p i t c h . P i t c h e s of between 1 a n d 2 are t y p i c a l . T h e m a i n advantage of he l i ca l s c a n n i n g is greater vo lume coverage per g a n t r y r o t a t i o n a n d therefore faster scanning . A d r a w b a c k is tha t a slice is not as we l l def ined as i n convent iona l C T a n d i n t e r p o l a t i o n is r e q u i r e d to reconstruct image slices. However , the fact t h a t scanned slice does not equa l reconstructed slice is a n advantage since slices can be reconstructed i n a n y a x i a l p lane of interest w i t h equal s p a t i a l a ccuracy [47, 53]. T h e speed of vo lume scanning is increased further w i t h the i n t r o d u c t i o n of mul t i - s l i c e C T . A mul t i - s l i c e C T scanner is capable of co l l ec t ing m u l t i p l e pro jec t ions at each angular p o s i t i o n of the gantry. A w ide x - r a y b e a m t e r m e d a cone beam is used a n d the detector a r r a y is parce l l ed i n the z d i re c t i on as we l l as w i t h i n the image plane, as i n convent ional , single slice C T . 2, 4, 8 a n d 16 slice machines are m a r k e t e d today. T h e y are t y p i c a l l y operated i n he l i ca l mode , as descr ibed above, a n d the technique is ca l led mul t i - s l i c e , he l i ca l C T . Advantages of m u l t i - s l i c e i m a g i n g inc lude very fast vo lume coverage t h a t al lows freezing of o rgan m o t i o n a n d efficient use of x - r a y tube o u t p u t . D r a w b a c k s inc lude reduced geometric a n d dose efficiencies a n d artefacts due to b o t h fast couch m o t i o n a n d cone b e a m x - r a y de l ivery [47]. I n general the performance of m u l t i - s l i c e C T scanners is c omparab le to single slice machines w i t h the advantage of s ign i f i cant ly reduced scan t i m e [58]. H i P-3 H 4 > • P l = M - l + H 2 * • P 2 = H 3 + H 4 p 5 =M., + u 3 p 4 = H 2 + ^ 4 F i g u r e 2.6: A n i l l u s t r a t i o n of the goal of r e cons t ruc t i on i n x - r a y C T : to reconstruct the d i s t r i b u t i o n of / i t h r o u g h ^4) i n a n object g iven measured pro j e c t i on samples ( p i t h r o u g h p5). 2.3 Image Reconstruction A s i n t r o d u c e d i n sect ion 2.1.3, C T images are p r o d u c e d f r o m p r o j e c t i o n d a t a . E a c h pro j e c t i on is the s u m m a t i o n of the a t t e n u a t i o n coefficients i n the scanned object t h a t f a l l a long the ray l ines between source a n d detector . G i v e n in f in i t e l y s m a l l s a m p l i n g a long the ray l ine , a pro j e c t i on (p) is g iven , i n theory, b y the l ine in tegra l , equat i on 2.3. Image reconstruc t i on i n C T is the process of r e c o n s t r u c t i n g the s p a t i a l d i s t r i -b u t i o n of a t t e n u a t i o n coefficients f r o m measured pro jec t ions to f o r m a n image of the object . F i g u r e 2.6 provides a s imple i l l u s t r a t i o n of th is goal . G i v e n measurements of p ro j e c t i on d a t a (pi t h r o u g h ps ) , but no a d d i t i o n a l knowledge about the object , the object a t t enuat i on coefficients (m t h r o u g h ^4) m u s t be de termined . A descr ip -t i o n of a n i n t u i t i v e approach to image recons t ruc t i on , i t e ra t ive reconstruct ion , is p r o v i d e d i n sect ion 2.3.1. Sec t ion 2.3.2 describes the theory of backpro j e c t i on , a n -other approach to image reconstruct ion . Sec t i on 2.3.3 provides detai ls o n f i l tered backpro j e c t i on , the reconstruc t i on technique c o m m o n l y i m p l e m e n t e d i n c o m m e r c i a l C T scanners. It is i m p o r t a n t to note that rea l w o r l d p r o j e c t i o n d a t a as o b t a i n e d f r om a C T scanner deviates f r o m the idea l i zed s i t u a t i o n g iven b y equat i on 2.3. T h i s is 1 2 3 4 > 3 •> 7 2.5 2.5 2.5 2.5 > 5 - > 5 (a) (b) 1.5 1.5 3.5 3.5 1 2 3 4 - > 3 > 7 5 5 4 6 (c) (d) F i g u r e 2.7: A n example of i t e ra t ive image reconstruc t i on . due to poly-energet ic x - r a y spectra , p h o t o n s ca t ter ing a n d detector n o n - l i n e a r i t y [47]. I n a d d i t i o n , pro ject ions depend o n scan geometry. F o r 1st a n d 2 n d generat ion scanners, as descr ibed i n figure 2.1a a n d b, at each angular scan p o s i t i o n r a y l ines are considered para l l e l a n d w h a t is t e r m e d a parallel projection is ob ta ined . However , for 3 r d a n d 4 t h generat ion scanners, s h o w n i n figures 2.1c a n d d, each angular scan p o s i t i o n produces a fan beam projection. F o r s imp l i c i t y , the r e cons t ruc t i on a l g o r i t h m detai ls presented here are for idea l ized , p a r a l l e l p ro j e c t i on d a t a . 2.3.1 I t e r a t i v e R e c o n s t r u c t i o n O n e so lut i on to image recons t ruc t i on is a n i t e ra t ive m e t h o d ca l led the a lgebraic r e cons t ruc t i on technique ( A R T ) . A s imple example of A R T is p r o v i d e d i n figure 2.7 a n d the approach is as fol lows. T h e object is i n i t i a l l y assumed to be u n i f o r m a n d a l l a t t enuat i on coefficients are set to prov ide the average of the pro j e c t i on samples . I n the example , figure 2.7a, each s u m m e d pro j e c t i on is 10 (7+3 or 4+6) . A v e r a g e d over a l l four p ixels each p i x e l is assigned a value of 2.5, as i n figure 2.7b. A pro j e c t i on is t h e n ca l cu la ted for th is u n i f o r m object . These values are c o m p a r e d w i t h the measured pro jec t ions a n d any difference d i v i d e d u n i f o r m l y amongst the p ixe ls a long each ray l ine . I n the example , the ca l cu la ted pro j e c t i on samples are 5. C o m p a r e d to the measured values of 3 a n d 7 the top row is overest imated b y 2 a n d the b o t t o m row u n d e r e s t i m a t e d by 2. A s a resul t , the t op row p ixe ls are reduced by 1 each a n d the b o t t o m row pixels increased b y 1 each, as shown i n figure 2.7c. T h i s process is repeated us ing other p r o j e c t i o n angles, u n t i l ca l cu la ted a n d measured pro ject ions agree. I n the example this is achieved i n one more i t e r a t i o n us ing the ver t i ca l p ro j e c t i on samples , as is s h o w n i n figure 2.7d. 2.3.2 B a c k p r o j e c t i o n A l t h o u g h A R T , as exempl i f ied b y f igure 2.7, is a successful reconstruc t i on technique , f o rward pro ject ions must be re - ca l cu la ted w i t h every i t e r a t i o n . T h i s makes the tech -n ique too c o m p u t a t i o n a l l y intensive to be p r a c t i c a l for fast, m o d e r n C T scanners. T h e i n i t i a l a s s u m p t i o n used i n A R T , tha t object in tens i ty is u n i f o r m a long a n x -r a y p a t h (as shown i n figure 2.7b), does lead to another i m p o r t a n t a p p r o a c h to image reconstruct ion , backpro j e c t i on . T h e fo l l owing p a r a g r a p h presents a n i n t u -i t ive discussion of b a c k p r o j e c t i o n a n d the c l i n i c a l a p p l i c a t i o n of th i s t o o l i n f i l tered b a c k p r o j e c t i o n a lgor th ims is de ta i l ed i n the f o l l owing sect ion, 2.3.3. T h e concept of b a c k p r o j e c t i o n is quite s imple . E v e r y p i x e l a long a ray l ine is g iven the in tens i ty of the recorded pro j e c t i on for tha t ray l ine . I n o therwords , each pro j e c t i on measurement is backprojected across the r a y l ine f r om w h i c h i t was co l lected a n d the backpro jec ted in tens i ty is s u m m e d over a l l pro ject ions . T h i s p r o -cess is repeated for each p r o j e c t i o n angle a r o u n d the object . F i g u r e 2.8 i l lus trates backprojeetions object image (a) (b) F i g u r e 2.8: A n i l l u s t r a t i o n of backpro j e c t i on image reconstruc t i on technique . T h e object , a u n i f o r m dot (a), is reconstructed b y b a c k p r o j e c t i n g the in tens i ty of mea -sured pro ject ions across the ray l ines (b). th i s backpro j e c t i on technique. P r o j e c t i o n s of the object , a u n i f o r m dot , are back-pro jec ted to o b t a i n a n image of the object . T h e object becomes v is ib le i n the image even for a re la t ive ly low number of backpro jeet ions , as is demonstra ted i n figure 2.8b. However , the image is a b l u r r e d vers ion of the object . T h i s is a n inherent feature of the backpro j e c t i on process t h a t remains even w h e n m a n y more pro jec -t ions are used [47]. A s descr ibed i n the fo l lowing sect ion, th is r educ t i on i n s p a t i a l reso lut ion is addressed by f i l ter ing the pro j e c t i on d a t a . 2.3.3 F i l t e r e d B a c k p r o j e c t i o n F i l t e r e d backpro j e c t i on is the most c o m m o n l y used m e t h o d of image recons t ruc t i on i n C T . T h e f u n d a m e n t a l pr inc ip le of backpro j e c t i on remains however the pro ject ions are f i ltered to remove the b l u r r i n g effect. Hence the name filtered backprojection. T h e basic goal of the filtering is to deconvolve the deleterious effect of the image f o r m i n g opera t i on (the s imple backpro jec t ion) f r o m the image, r e t u r n i n g the o r i g i n a l object . I n pract i ce filtered b a c k p r o j e c t i o n is often per formed i n the Four i e r d o m a i n to avo id the convo lut i on operat ions . I n what fol lows, the a l g o r i t h m a n d filters are descr ibed i n more de ta i l . A l g o r i t h m T h e f i l tered backpro j e c t i on a l g o r i t h m opera t ing i n Four i e r space (v, v) employs the Four i e r slice theorem [47]. T h i s theorem states tha t the I D Four i e r t r a n s f o r m ( F T ) of a pro j e c t i on of a n object at a g iven angle equals a l ine i n the 2 D F T of the object t a k e n at the same angle. T h i s is best i l l u s t r a t e d b y a n example . Cons ider a pro j e c t i on of the object f(x,y) para l l e l to the y -ax is , p(x,0). B y de f in i t i on , /oo f(x,y)dy. (2.5) - O O T h e F T of p(x, 0) gives /oo roo / f(x,y)e-2™xdxdy. (2.6) •oo J—oo T h e r ight h a n d side of th is equat i on is the 2 D F T of f{x,y), P(V,L>), eva luated for v = 0. I n other words, the F T of the 0° pro j e c t i on is the same as the v = 0 l ine i n the 2 D F T of the same object . T h u s , a n image of a n object c a n be reconstructed d i rec t ly f r o m Four ier space b y a n inverse 2 D F T . T h i s is however more di f f icult to i m p l e m e n t t h a n i t m i g h t appear due to the need to resample the po lar d a t a ob ta ined i n Four i e r space in to cartes ian coordinates necessary to represent the image [47]. T h e f i l tered b a c k p r o j e c t i o n a l g o r i t h m proceeds as follows. F i l t e r s are a p -p l i ed , v i a m u l t i p l i c a t i o n s , to the F T of the pro j e c t i on d a t a . T h e n , based o n the Four i e r slice theorem, the object is reconstructed us ing inverse F T operat ions . I n i m p l e m e n t a t i o n , a po lar coord inate sys tem (u, 0) is used for the p r o j e c t i o n d a t a i n Four i e r space. G i v e n the F T of a pro j e c t i on is P(u),9), the f u n d a m e n t a l f i l tered b a c k p r o j e c t i o n equat i on is rir roo f{x,y)= d0 P(uj,e)\Lj\e-2^xcose+ysin^dw. (2.7) J0 J-oo N o t e t h a t the in tegra t i on l i m i t s i n 9 der ive f r om s y m m e t r y i n the p r o j e c t i o n d a t a [47]. T h e fi lter used here is |w| a n d is ca l led a ramp filter. I n t u i t i v e l y , use of a r a m p fi lter makes sense. I n Four i e r space a l l pro jec t ions w i l l cross the o r i g i n . M u l t i p l i c a t i o n b y frequency w i l l correct the low frequency we ight ing a n d the b l u r r i n g inherent i n backpro jec t i on . However , a r a m p fi lter w i l l restore edges a n d noise due to the increased we ight ing of h i g h frequency image components . F i l t e r design is a n i m p o r t a n t m e t h o d of m o d i f y i n g noise character ist i cs i n the reconstructed image a n d f i lters are discussed i n more d e t a i l below. F i l t e r s T h e backpro j e c t i on f i lter as descr ibed above operates over a l l freqencies ( f rom — oo t o oo). R e a l w o r l d fi lters are b a n d - l i m i t e d . A fi l ter , H(u>): is descr ibed b y H(u) = \u>\q{u) (2.8) where q(u) is a w i n d o w used to b a n d - l i m i t the f i lter . F o r a r a m p f i l ter , a s imple square w i n d o w is used. M o d i f i c a t i o n s to the w i n d o w f u n c t i o n change the frequency response of the fi lter a n d therefore alter i ts effect o n image noise a n d s p a t i a l reso lu -t i o n . A n example is use of a sine f u n c t i o n w i n d o w , I I s ( w ) , def ined b y , 1, M < w/, T\a(uj) = } s i n [ T T ( | O J I ) / ( u H - u L ) } , ^ I , ,1 < . , „ f O Q~) nS(uj) - < n(M_UJL)/(UH_U)L) , W L < \UJ\ ^ UJH l z - y J 0, > OJH where LOH a n d UJL are the u p p e r a n d lower l i m i t s o f the frequencies t h a t w i l l be modi f i ed , respectively. C o m p a r e d to the r a m p fi lter th i s f i lter provides greater noise r e d u c t i o n b y preferent ia l ly suppress ing h i g h frequencies. I n a d d i t i o n , a l t e r i n g the cut-of f frequency w i l l i m p a c t image noise a n d s p a t i a l r eso lut i on propert ies . F o r example , a v is ib le r educ t i on i n noise level is achieved b y r e d u c i n g CJH i n the sine f i l ter [47]. T y p i c a l l y , m o d e r n C T scanners are equ ipped w i t h several different prepro -g r a m m e d reconstruc t i on opt ions t h a t can be selected de pe n d in g o n the requirements of each app l i ca t i on . F o r example , G E C T scanners have s ix different re cons t ruc t i on a lgor i thms cal led S O F T , S T A N D A R D , D E T A I L , L U N G , B O N E a n d E D G E [54]. It is the fi lters employed i n these a n d other c o m m e r c i a l r e cons t ruc t i on a l g o r i t h m s t h a t provide the v a r y i n g effects o n the reconstruc ted image. T h e design of these fi lters is a c o m p l i c a t e d process a n d the i r detai ls are p r o p r i e t a r y a n d not general ly released i n the p u b l i c d o m a i n . 2.4 Noise and Artefacts I n a d d i t i o n to reconstruc t i on artefacts , C T images are degraded b y image noise a n d other types of image artefacts. H i g h levels of noise c a n obscure objects i n a n image, p a r t i c u l a r l y i n low contrast regions [47]. F o r example , low image noise is very i m p o r t a n t for detect ion of a b r a i n t u m o u r o n a C T image. A n image artefact is any inaccura te ly represented NQT • A r t e f a c t s are t roublesome i n two ways. T h e y can obscure the a c t u a l image of the ob ject a n d t h e y can also be m i s i n t e r p r e t e d as rea l d a t a . T h e r e d u c t i o n of b o t h image noise a n d artefacts are i m p o r t a n t for C T gel dos imetry . Noise a n d artefacts i n C T images are discussed here i n sections 2.4.1 a n d 2.4.2, respectively. 2.4.1 N o i s e T h e r e are three p r i m a r y sources of noise i n C T images. T h e most s ignif icant is q u a n t u m or s ta t i s t i ca l noise. Q u a n t u m noise is inherent i n p h o t o n detec t ion a n d results f r o m c o u n t i n g a f inite n u m b e r of r a n d o m events. Q u a n t u m noise is P o i s s o n d i s t r i b u t e d a n d the uncer ta in ty i n NQT , O~NCti due *° q u a n t u m noise is g iven by where N is the n u m b e r of photons detected. A s is i n d i c a t e d b y th i s equat ion , <T ]v c r is reduced b y increas ing N. A s descr ibed i n sect ion 2.2.3, choice of scan p r o t o c o l can affect N a n d this is the most effective means a scanner operator has of r e d u c i n g image q u a n t u m noise. I n a d d i t i o n , the na ture of the scanned object i tsel f w i l l affect N a n d therefore the amount of q u a n t u m noise degrad ing the image. M a t e r i a l density, 1 (2.10) a t o m i c number a n d size a l l affect image noise t h r o u g h increas ing or decreasing x - r a y a t t enuat i on i n the object . T h e other m a j o r sources of noise i n C T images are p h y s i c a l l i m i t a t i o n s of the system a n d the image generat ion process. P h y s i c a l l i m i t a t i o n s are e lectronic noise i n the detectors a n d d a t a a c q u i s i t i o n systems. I n a d d i t i o n , the geometry of the scanner, speci f ical ly the source to detector d is tance , w i l l affect N a n d therefore image q u a n t u m noise. T h e operator cannot reduce these sources of image noise. W i t h i n the image generat ion process, as was discussed i n sect ion 2.3.3, f i lters used i n image reconstruct ion affect r e s u l t i n g image noise levels. A s a result , select ion of reconstruct ion a l g o r i t h m is another means a scanner operator has to reduce image noise. Noise r educ t i on is i m p o r t a n t for C T gel dos imetry a n d is discussed further i n chapter 6. I n a d d i t i o n , the p o t e n t i a l of post -process ing gel C T images to further reduce image noise is discussed i n chapter 7. 2.4.2 A r t e f a c t s Alongs ide image noise, artefacts are another signif icant source of image degradat ion . T h e r e are m a n y p o t e n t i a l artefacts i n C T i m a g i n g [47] a n d they c a n take several forms i n c l u d i n g streaks, s h a d i n g a n d r ings . T h e y are due, i n c o m b i n a t i o n w i t h the image reconstruct ion a l g o r i t h m , to a var ie ty of factors re la ted to sys tem design, x -r a y tube performance , detector per formance a n d the object or pat ient be ing imaged . T h e fo l lowing describes some of the m o r e c o m m o n l y observed C T i m a g i n g artefacts. O n e c o m m o n artefact is the b e a m h a r d e n i n g artefact . I n polyenerget ic x -r a y beams, such as those f ou n d i n C T scanners, l ow energies are pre ferent ia l ly a t tenuated as the b e a m passes t h r o u g h the object . T h i s is t e r m e d beam hardening. I n a C T image th is can produce a cupping artefact w h i c h shows a n enhancement of s igna l intens i ty at the edges of h i g h dens i ty objects . Ob je c t s w i t h very h i g h dens i ty compared to s u r r o u n d i n g mater ia l s , such as m e t a l i m p l a n t s i n t issue, can exh ib i t c ompl i ca ted art i facts t h a t are due to b e a m h a r d e n i n g c o m b i n e d w i t h p a r t i a l vo lume a n d under range d a t a effects. T h e p a r t i a l vo lume effect is due to a h i g h densi ty m a t e r i a l o c c u p y i n g o n l y par t of the vo lume traversed by the x - r a y b e a m . T h i s results i n a shadow l ike streak i n the image. Use o f s m a l l sl ice thickness reduces th i s type of artefact . U n d e r range d a t a artefacts also appear as streaks a n d are the result of the d a t a be ing under range i n the d a t a a c q u i s i t i o n sys tem due to extreme p h o t o n a t t e n u a t i o n b y the h i g h dens i ty m a t e r i a l . Ar te fac t s t h a t appear as r ings are also c o m m o n i n C T i m a g i n g . T h e y can be produced i n several ways , a l l re la ted to detector per formance . Detec tor fa i lure or m i s c a l i b r a t i o n i n a t h i r d generat ion, ro tate - ro tate scanner, w i l l produce a r i n g artefact where r i n g rad ius relates to the p o s i t i o n i n the detector a r ray of the fau l ty detector. One character i s t i c of these r i n g artefacts is t h a t artefact in tens i ty increases as r i n g rad ius decreases [47]. Detec tors w i t h n o n - u n i f o r m response w i l l also produce r i n g artefacts. T h i s is f requent ly a s y m p t o m of r a d i a t i o n damage a n d is a c o m m o n p r o b l e m as detectors age. I n some of these cases, artefacts can be removed or at least reduced i n i n t e n -s i ty by f i x ing the p r o b l e m at the root of a n artefact . F o r example , a fau l ty detector c o u l d be replaced or re - ca l ibra ted . I n contrast , where the artefact results f r o m the na ture of the scanned object itself, the root p r o b l e m cannot be fixed. I n these cases i t is possible to remove some artefacts b y s u b t r a c t i n g a b a c k g r o u n d image. A r t e f a c t removal is i m p o r t a n t for C T gel d o s i m e t r y as image contrast due to the presence of artefacts m a y be mis - in terpre ted as dose i n f o r m a t i o n . A s is discussed further i n chapters 4 a n d 5, b a c k g r o u n d s u b t r a c t i o n is f requent ly employed i n C T gel dos imetry to remove artefacts . Chapter 3 Polymer Gel Dosimetry T h i s chapter provides a n i n t r o d u c t i o n to the f ield of gel dos imetry , i n p a r t i c u l a r to p o l y m e r gel dos imetry . A n overv iew of gel dos imetry i n general is p r o v i d e d i n sect ion 3.1. A more deta i l ed l ook at the t y p e of gel invest igated i n th i s thesis , p o l y a c r y l a m i d e gel or P A G , fol lows i n sect ion 3.2. Sec t i on 3.3 prov ides a n overview of current l i t e rature i n p o l y m e r gel dos imetry . T h i s is fol lowed b y a de ta i l ed d iscuss ion on x - r a y c o m p u t e d t o m o g r a p h y for r e a d i n g out gel dosimeters ( C T gel dos imetry ) i n sect ion 3.4, as this t op i c is the focus of th i s thesis . F i n a l l y , sect ion 3.5 discusses the concept of dose reso lut ion as a p p l i e d to C T gel dos imetry . 3.1 Introduction 3.1.1 A B r i e f H i s t o r y o f G e l D o s i m e t r y T h e idea of us ing gels to measure dose was explored as ear ly as the 1950s. Changes i n co lour a n d spec t rophotometry measurements were used to analyse r a d i a t i o n i n d u c e d changes i n gels [59, 60]. However , i t was the i n t r o d u c t i o n i n 1984 of a Fr i cke gel dosimeter b y J . G o r e a n d colleagues at Y a l e U n i v e r s i t y t h a t s ta r t ed the m o d e r n f ield of gel dos imetry [30, 61]. T h i s ear ly gel system consisted of the we l l establ ished Fr i cke chemica l dosimeter [29] h in fused i n a ge la t in m a t r i x i n order to record s p a t i a l dose i n f o r m a t i o n . T h e r a d i a t i o n i n d u c e d changes i n Fr i cke s o l u t i o n ( F e 2 + —• F e 3 + ions) p rov ide corresponding changes i n nuc lear magnet i c resonance propert ies a n d magnet i c resonance i m a g i n g was used to ex trac t dose i n f o r m a t i o n f r o m the gels [30, 61]. U n f o r t u n a t e l y , due to the i r s m a l l size, ferrous ions can diffuse t h r o u g h the gel p r o d u c i n g s p a t i a l i n s t a b i l i t y i n these dosimeters [62-64]. However , Fr i cke gel d o s i m e t r y remains a n act ive b r a n c h of gel d o s i m e t r y research t o d a y [65], p a r t i c u l a r l y w i t h the a d d i t i o n of co lour dyes a n d a p p l i c a t i o n of new o p t i c a l s cann ing techniques (see sect ion 3.3.2). 1992 saw the next m a j o r step i n gel d o s i m e t r y w i t h the i n t r o d u c t i o n of a new t y p e of system based on r a d i a t i o n i n d u c e d p o l y m e r i z a t i o n of monomers [31]. T h i s f irst polymer gel consisted of the monomers a c r y l a m i d e a n d N, ./V"'methylene b i s a c r y l a m i d e , in fused i n a n agarose gel m a t r i x . These molecules c ross - l ink a n d po ly me r i ze u p o n i r r a d i a t i o n to f o r m a b r a n c h i n g p o l y m e r network . D e t a i l s o n these molecules a n d p o l y m e r i z a t i o n processes are p r o v i d e d i n sect ion 3.2. A s w i t h the Fr i cke gels, M R I proved a useful t o o l for r ead ing out the s p a t i a l dose i n f o r m a t i o n as the N M R transverse r e l a x a t i o n rate ( R 2 ) of the gel increases w i t h p o l y m e r i z a t i o n a n d , hence, dose del ivered to the gel. T h i s new p o l y m e r based sys tem was successful i n overcoming the di f fusion l i m i t a t i o n of F r i c k e gels due to the large size of the f o rmed p o l y m e r molecules [31]. A f t e r th is i n i t i a l i n t r o d u c t i o n of p o l y m e r gel , the agarose was rep laced by ge la t in to create the B i s A c r y l a m i d e N i t r o g e n G e l a t i n ( B A N G ™ ) gel [66] a n d the first c l i n i c a l app l i cat ions were d e m o n s t r a t e d [67-69]. B A N G r M is now t r a d e m a r k e d a n d is reserved for gels purchased f r o m M G S Inc. ( M a d i s o n C T , U S A ) a n d the terms P A G or, less frequently, B A N G - t y p e , are used instead . P A G remains the backbone of p o l y m e r gel dos imetry t o d a y a n d i t is descr ibed i n more d e t a i l i n sect ion 3.2. Since th is t i m e , research i n gel dos imetry has focused o n f u n d a m e n t a l w o r k u n d e r s t a n d i n g , charac ter i z ing a n d developing p o l y m e r gels, deve loping a l ternate i m a g i n g moda l i t i e s a n d inves t igat ing c l i n i c a l app l i cat ions . Deve lopments i n these areas are discussed i n the review of current l i t e ra ture , sect ion 3.3. 3.1.2 O v e r v i e w o f t h e T e c h n i q u e F o r a l l gel dosimeters , the overal l process for o b t a i n i n g a 3 D dose measurement is the same. T h e basic steps are shown schemat i ca l ly i n f igure 3.1. T h e gel is f irst m a n u f a c t u r e d f r o m i ts const i tuent components , f igure 3.1a. F o r P A G th is general ly requires use of a glove box, as descr ibed i n chapter 4. It s h o u l d be no ted t h a t , as ment i oned above, i t is cur rent ly possible to o m i t t h i s step as some types of p o l y m e r gel are avai lable for purchase f r o m M G S Inc. . A f t e r manufac ture , the gel is i r r a d i a t e d w i t h the t reatment for w h i c h a 3 D dose d i s t r i b u t i o n is desired, figure 3.1b. W h e n p o l y m e r i z a t i o n is complete , the gel is i m a g e d to ex trac t the recorded dose d i s t r i b u t i o n , figure 3.1c. T h i s is t r a d i t i o n a l l y done u s i n g M R I , b u t , as descr ibed i n sect ion 3.3.2, m u c h current gel dos imetry research is focused on development of a l ternate i m a g i n g moda l i t i e s for gel read-out . A n e x a m p l e is x - r a y C T , the focus of th is thesis. F o r 3 D dos imetry , the i m a g i n g step m u s t image the entire 3 D gel vo lume, r e q u i r i n g a c q u i s i t i o n of a series of slices t h r o u g h the dosimeter . T h e f ina l step is a processing step i n order to convert gel images to dose maps , figure 3 . Id . T h i s m a y s i m p l y require n o r m a l i z a t i o n w h e n p e r f o r m i n g re lat ive dos imetry g iven a l inear dose response, or require the more c o m p l i c a t e d a p p l i c a t i o n of a c a l i b r a t i o n curve w h e n per f o rming absolute dose measurement or u s i n g a non- l inear dose response. A d d i t i o n a l process ing steps (e.g. image reg istrat ion) are r e q u i r e d i f the f ina l goal is c ompar i son of measured a n d p l a n n e d dose d i s t r i b u t i o n s . 3.1.3 A d v a n t a g e s o f G e l D o s i m e t r y Des i rab le character ist i cs for r a d i a t i o n dosimeters were s u m m a r i z e d i n sect ion 1.3.1. T h e advantages of gel d o s i m e t r y s t e m f rom the fact t h a t i t has several of these char -acterist ics not present i n other dosimeters c o m m o n l y avai lable i n r a d i a t i o n t h e r a p y (i.e. i o n chambers , diodes, T L D s a n d film). T h e first is the a b i l i t y to measure r a d i -(a) (b) m > I — _ I s ^ (c) (d) F i g u r e 3.1: A s impl i f i ed overview of a s t a n d a r d gel dos imetry process. T h e gel is m a n u f a c t u r e d (a) a n d t h e n i r r a d i a t e d w i t h a t reatment for w h i c h a measured dose d i s t r i b u t i o n is desired (b). T h e p o l y m e r i z e d gel is t h e n i m a g e d to extract the dose i n f o r m a t i o n (c) a n d the images are processed to produce dose maps (d). a t i o n dose i n 3 D w i t h very h i g h s p a t i a l reso lut ion . T h e gels themselves c a n record dose to w i t h i n micrometers a n d i t is the gel i m a g i n g systems w h i c h l i m i t achievable reso lut i on to, typ i ca l ly , s u b - m i l l i m e t e r or m i l l i m e t e r scales. A s such , i f p r o b e d w i t h a very h i g h reso lut ion i m a g i n g m o d a l i t y , m e a s u r i n g dose d i s t r i b u t i o n s w i t h m i c r o m -eter prec i s ion w o u l d be possible [70]. A second advantage of gel d o s i m e t r y is the rad io l og i ca l t issue equivalence of p o l y m e r gel [71, 72]. T h i s is i m p o r t a n t as i t means t h a t a measured dose to the gel is equivalent to the dose to t issue or water , the s t a n -d a r d m a t e r i a l considered to be rad io l og i ca l l y t issue equivalent . O n e final advantage is the a b i l i t y of p o l y m e r gels to be m o l d e d into a n t h r o p o m o r p h i c p h a n t o m s . T h i s un ique a b i l i t y is due to the t r a n s i t i o n of the gel mater ia l s f r o m a l i q u i d to a gel state d u r i n g m a n u f a c t u r i n g . A p o l y m e r gel can be p o u r e d in to a p h a n t o m w h i c h m i m i c s ex terna l a n a t o m i c a l contours a n d / o r incorporates inhomogenei t ies such as a i r or bone before set to gel so l id [73-75]. 3.2 Polyacrylamide Gels A s descr ibed i n sect ion 3.1.1, p o l y a c r y l a m i d e gel , t e r m e d P A G , is the backbone of rad i o therapy p o l y m e r gel dos imetry a n d is the t y p e of p o l y m e r gel invest igated i n th is thesis. T h i s sect ion provides a n overv iew of P A G c o m p o s i t i o n , sect ion 3.2.1, p o l y m e r i z a t i o n react ions , sect ion 3.2.2, a n d basic dose response character is t i cs , sec-t i o n 3.2.3. 3.2.1 C o m p o s i t i o n P A G c o m p o s i t i o n is re lat ive ly s imple , cons is t ing of a c r y l a m i d e , N, N' methy lene b i s a c r y l a m i d e , ge la t in a n d water . T h e act ive molecules are the a c r y l a m i d e a n d N, A ' ' 'methylene b i sac ry lamide , w h i c h , u p o n i r r a d i a t i o n , f o r m cross - l inked p o l y -a c r y l a m i d e . A c r y l a m i d e is a short , l inear monomer w i t h the c h e m i c a l f o r m u l a CH2 : CHCONH2 a n d a chemica l s t ruc ture as shown i n figure 3.2a. T h e chemica l s t ruc ture of A'", N methylene b i s a c r y l a m i d e , c o m m o n l y k n o w n as bis, ( chemica l for-m u l a : (CH2 • CHCONH)2 - CH2) is shown i n figure 3.2b. N o t i c e t h a t the l inear a c ry lamide monomer has one c a r b o n double b o n d , whereas bis has two. These two c a r b o n double bonds present i n bis a l l ow it to f o r m branched molecules o n p o l y -m e r i z a t i o n and , as such, bis is t e r m e d a crosslinker. T h i s c ross l ink ing is evident i n the chemica l s t ruc ture of c ross - l inked p o l y a c r y l a m i d e , f o rmed o n c o - p o l y m e r i z a t i o n of a c ry lamide a n d b is , as shown i n figure 3.2c. T h e role of the ge la t in i n P A G is to gel the system, p r o v i d i n g a s t u r d y network t h a t ensures the s p a t i a l in tegr i ty of the dosimeter . A s t a n d a r d P A G c o m p o s i t i o n is g iven i n tab le 3.1. N o t i c e t h a t the gel is near ly 9 0 % water . T h i s accounts large ly for i ts t issue equivalence a n d makes the i n t e r a c t i o n of r a d i a t i o n w i t h water the p r i m a r y m e c h a n i s m for r a d i c a l p r o d u c t i o n a n d i n i t i a t i o n of p o l y m e r i z a t i o n . T h i s process is d iscussed i n some de ta i l i n the fo l lowing sect ion, 3.2.2. T h e c o m p o s i t i o n of c ross - l inked p o l y m e r gels is f requent ly descr ibed i n terms of the t o t a l a m o u n t of m o n o m e r i n the gel (%T) a n d the re lat ive f rac t i on of c ross - l ink ing molecule to t o t a l m o n o m e r ( % C ) . F o c u s i n g on P A G i n p a r t i c u l a r , % T is the weight f r a c t i o n of a c r y l a m i d e a n d bis combined , a n d % C is the weight f rac t i on of b is re lat ive to tha t of a c r y l a m i d e . I n other words , i f the weight fract ions of a c r y l a m i d e a n d bis are A a n d B %, respect ively , t h e n %T = A + B (3.1) a n d %C = —?— x 100% (3.2) A + B v ; U s i n g these def init ions the s t a n d a r d P A G f o r m u l a t i o n , g iven i n tab le 3.1, is descr ibed as 6 % T , 50 % C . % T a n d % C are used to define the var ious P A G formulat ions invest igated i n the c o m p o s i t i o n a l s tudy , chapter 5. 3.2.2 P o l y m e r i z a t i o n T h i s sect ion describes the r a d i a t i o n react ions a n d p o l y m e r i z a t i o n of P A G . Br ie f ly , i r r a d i a t i o n of P A G produces water radica ls w h i c h react w i t h gel monomers to p r o -C H , ^CH I G = Q I X H , (a) C H = CH, 1 c = o I NH I C H 2 I N H CI I, -c = o CI I (b) I N H , C I I , - C H -i c o I N H , CI I 1 "CIL - C H -i C H , 1 ; c ••= O 1 C = 0 1 NH" \ C H 2 [ N H : i J h l M l ' i 1 C -1 O 1 CH - C H , C H - C H 2 -1 c - 0 j N H 2 : _ n - C H . CU I ' c:= o 1 NH 1 C H 2 I NH •I" C = O - C H 2 - CH -(c) F i g u r e 3.2: T h e chemica l s t ructure of a c r y l a m i d e (a), ./V, ^ ' m e t h y l e n e b i s a c r y l a m i d e (b) a n d the i r r a d i a t i o n produc t , c ross - l inked p o l y a c r y l a m i d e (c). N o t e the b r a n c h e d nature of the cross - l inked p o l y a c r y l a m i d e molecule . T a b l e 3.1: T h e s t a n d a r d P A G f o r m u l a t i o n . C o m p o n e n t % b y weight water 89 ge la t in 5 a c r y l a m i d e 3 bis 3 duce act ive monomer rad i ca l s . These m o n o m e r radica ls propagate p o l y m e r i z a t i o n a n d p o l y m e r is p r o d u c e d i n a n amount p r o p o r t i o n a l , i n some way, to the number of water radica ls p r o d u c e d a n d hence to the dose absorbed l o c a l l y i n the gel. T h e fo l lowing provides detai ls of these processes. W a t e r radiolysis A s descr ibed i n the prev ious sect ion, P A G is a p p r o x i m a t e l y 90 % water b y weight . A s such, u p o n i r r a d i a t i o n , dose is absorbed p r i m a r i l y b y water . T h e in te rac t i on of i o n i z i n g r a d i a t i o n w i t h water , t e r m e d water radiolysis, is a w e l l s t u d i e d f ield a n d o n l y a br ie f d iscuss ion is p r o v i d e d here. T h e reader is referred to s t a n d a r d r a d i a t i o n chemis t ry t ex tbooks for greater d e t a i l (e.g. t h a t b y A . J . S w a l l o w [76]). T h e r e are three stages of water rad io lys i s . F i r s t , water molecules are i on i zed or exc i ted to produce either a water c a t i o n (H20+) a n d a n e lectron (e~) or a n exc i ted water molecule (H20*): H20 H20+ + e~,H20* (3.3) T h e second stage sees these three species undergo ing further react ions to produce h y d r o x y l a n d hydrogen rad ica l s (OH* a n d H*, respectively, where • represents a n u n p a i r e d electron) , h y d r o n i u m ions (H%0+) a n d h y d r a t e d electrons (e~ g) as follows: H20+ + H20 - » # 3 0 + + OH* (3.4) e - + n i f r O - > e~ (3.5) H20* - » H • +OH* (3.6) I n the f ina l stage of water rad io lys i s , these p r i m a r y reac t i on produc ts {e~q, OH», H* a n d H^O+) react w i t h each other , water molecules , or solutes present i n the system, to es tab l i sh chemica l e q u i l i b r i u m . I.e.,: ^ + e " + 2H20 - H2 + 20H~ (3.7) e~ + H30+ - H . +H20 (3.8) e~q + H* +H20 ^H2 + OH' (3.9) 2H* i ? 2 (3.10) 20H» H202 (3.11) e " + O f f . - v O f f " (3.12) H*+OH»^H20 (3.13) e~q + H20 ^ H »+OH- (3.14) e~q + H202 -+OH »+OH- (3.15) OH • +H2 H20 + H* (3.16) These basic react ions w i l l occur i n i r r a d i a t e d P A G , however i t is the r eac t i on of these p r i m a r y water radica ls w i t h monomers i n P A G t h a t in i t ia tes p o l y m e r i z a t i o n , as descr ibed below. R a d i c a l c h a i n p o l y m e r i z a t i o n P o l y m e r i z a t i o n is the process b y w h i c h monomers j o i n to f o r m po lymer . W h e n two different monomers are invo lved , the process is correc t ly t e r m e d c o - p o l y m e r i z a t i o n , however i n th i s d iscuss ion the t e r m p o l y m e r i z a t i o n is used for s impl i c i ty . T h e r e are two m a i n types of p o l y m e r i z a t i o n : step p o l y m e r i z a t i o n a n d cha in p o l y m e r i z a t i o n . P A G gel po lymer izes b y cha in p o l y m e r i z a t i o n , i n p a r t i c u l a r , b y radical chain poly-merization [77]. T h i s t e r m refers to the fact t h a t radica ls i n i t i a t e the p o l y m e r i z a t i o n react ions. C a t i o n s , anions or the i n t r o d u c t i o n of a t rue cata lys t are a d d i t i o n a l means of i n i t i a t i n g c h a i n p o l y m e r i z a t i o n [78]. R a d i c a l c h a i n p o l y m e r i z a t i o n proceeds i n three steps: initiation, propagation a n d termination. In i n i t i a t i o n , a n act ive m o n o m e r is p roduced as descr ibed by : / • + M - > A f « (3.17) where / • is the r a d i c a l i n i t i a t o r a n d M a n d M « are inact ive a n d act ive monomers , respectively. I n P A G water rad i ca l s f o rmed b y i r r a d i a t i o n , as descr ibed above, are the r a d i c a l in i t i a t o r s . N o t e t h a t the presence of oxygen i n h i b i t s the i n i t i a t i o n of p o l y m e r i z a t i o n a n d P A G must be free of oxygen i n order for p o l y m e r i z a t i o n to occur . R a d i c a l c h a i n p o l y m e r i z a t i o n propagates b y act ive monomers (or m o n o m e r chains) reac t ing w i t h single monomers to p roduce act ive m o n o m e r chains increased i n l eng th by one monomer . T h i s is represented by : Mn»+M ^ M n + 1 . (3.18) where the subscr ipts are the n u m b e r of monomers i n the cha in . C h a i n l ength growing by one monomer at a t i m e is a character i s t i c of c h a i n p o l y m e r i z a t i o n . P o l y m e r i z a t i o n t e rminates i n the p r o d u c t i o n of a n inact ive or dead p o l y m e r . T h i s can occur b y processes t e r m e d recombination a n d disporportionation. R e c o m -b i n a t i o n is the process b y w h i c h two act ive m o n o m e r chains b o n d to f o r m a n inac t i ve po lymer . It is descr ibed by : Mn • + M m . -> Mn+m (3.19) where subscr ipts , again , represent c h a i n l ength . D i s p o r p o r t i o n a t i o n is the process by w h i c h one act ive m o n o m e r c h a i n pu l l s a hydrogen i o n f r o m another act ive m o n o m e r cha in , p r o d u c i n g two dead po lymers , lengths equa l to the o r i g i n a l act ive chains . It is descr ibed by: Mn • + M m . -^Mn + Mm (3.20) 68 a n d is compet i t ive w i t h r e c o m b i n a t i o n i n t e r m i n a t i n g p o l y m e r i z a t i o n [78]. P A G po lymer izes b y this process of r a d i c a l c h a i n p o l y m e r i z a t i o n . T h e s i t u a -t i o n is however somewhat more c o m p l e x as there is a n a d d i t i o n a l component to the system, ge lat in . A decrease i n P A G dose response has been shown w i t h increased ge lat in concentrat i on for b o t h M R I a n d C T read-out [43, 79]. These results suggest t h a t ge la t in scavenges some of the rad ica ls , r e d u c i n g the n u m b e r avai lable to i n i t i a t e p o l y m e r i z a t i o n . 3.2.3 D o s e R e s p o n s e B r o a d l y speaking , the dose response of a s t a n d a r d P A G , such as the f o r m u l a t i o n g iven i n tab le 3.1, is h i g h at low doses a n d begins to saturate at h igher doses. It can t y p i c a l l y be wel l descr ibed by a m o n o - e x p o n e n t i a l func t i on , w i t h the low dose region often considered quasi-linear for easy i m p l e m e n t a t i o n of re lat ive dos imetry . A sketch of a classic P A G dose response is p i c t u r e d i n figure 3.3. T h i s character i s t i c response is observed w i t h a l l read-out m o d a l i t i e s (as descr ibed i n sect ion 3.3.2) a n d m i m i c s tha t seen for p o l y m e r f o r m a t i o n u s i n g R a m a n spectroscopy [80]. It m u s t be noted that this dose response is s h o w n here to i l l u s t r a t e the general way i n w h i c h P A G responds to dose. T h e r e are m a n y factors w h i c h affect P A G dose response (gel c ompos i t i on , t empera ture etc.) , some r e s u l t i n g i n d r a m a t i c changes to b o t h the sens i t iv i ty a n d f u n c t i o n a l f o rm. M a n y of these factors w i l l be discussed i n the l i t e ra ture review, sect ion 3.3. 3.3 Polymer Gel Dosimetry: A Literature Review G e l dos imetry is a n e x p a n d i n g f ie ld a n d i n recent years the l i t e ra ture p e r t a i n i n g to gel dos imetry has become qui te extensive. T h i s sect ion provides some of the h igh l ights i n three m a i n areas of gel d o s i m e t r y research: f u n d a m e n t a l studies on gel propert ies a n d the development of new gels, sect ion 3.3.1, development of gel i m a g i n g modal i t i es , sect ion 3.3.2, a n d inves t iga t i on of gel app l i ca t i ons , sect ion 3.3.3. A Dose • F i g u r e 3.3: A sketch i l l u s t r a t i n g the na ture of a s t a n d a r d P A G gel response to dose. T h i s classic response is exponent ia l w i t h a quasi-linear reg ion at low doses often used for s impl i f i ed re lat ive dos imetry . It is i m p o r t a n t to note t h a t there are m a n y factors w h i c h affect th is dose response a n d w h a t is shown here is a sample response for i l l u s t r a t i v e purposes. T h e reader is also referred to the proceedings of the three i n t e r n a t i o n a l conferences on rad io therapy gel dos imetry , D O S G E L 9 9 , D O S G E L 2 0 0 1 a n d D O S G E L 2 0 0 4 for good reviews of current research [65, 81, 82]. 3.3.1 F u n d a m e n t a l G e l S t u d i e s M u c h f u n d a m e n t a l w o r k i n gel dos imetry has focused o n s t u d y i n g factors t h a t af-fect P A G dose response a n d s t u d y i n g gel propert ies such as t e m p o r a l a n d s p a t i a l s tab i l i ty . Recent ly , spurred p a r t l y by the t o x i c i t y of a c r y l a m i d e [83] a n d the i n t r o -d u c t i o n of the first r a d i a t i o n dos imetry p o l y m e r gel t h a t c a n be made under n o r m a l a tmospher i c condi t ions [84], the development of new p o l y m e r gel f ormulat ions has emerged as another h i g h l y act ive area of f u n d a m e n t a l work . T h e fo l lowing h igh l ights some of the l i t e rature i n these areas. Factors Af fect ing D o s e R e s p o n s e A p a r t f r om the presence of oxygen , w h i c h even i n low concentrat ions can prevent p o l y m e r i z a t i o n [85], gel c o m p o s i t i o n has the most s ignif icant effect o n P A G dose response. T h e t o t a l c oncentra t i on of m o n o m e r (%T) a n d the cross l inker f r a c t i o n (%C) affect the sens i t iv i ty a n d f u n c t i o n a l f o rm ( in the case of % C ) of P A G dose response [35, 80, 86, 87]. U s i n g R a m a n spectroscopy, gel c o m p o s i t i o n has been shown to d i rec t ly affect the rates of m o n o m e r c o n s u m p t i o n a n d p o l y m e r f o r m a t i o n [80, 86, 87]. G e l a t i n concentra t i on also affects dose response. Increasing the amount of ge la t in decreases P A G sens i t i v i ty to b o t h M R I a n d C T read-out [43, 79, 88]. T h i s is l ike ly a result of the r a d i c a l scavenging a b i l i t y of ge la t in [79]. G e l i m a g i n g t emperature is another factor g rea t ly affecting the sens i t iv i ty of P A G M R I dose response [34, 35, 89]. DeDeene et al. have shown t h a t changes i n gel t empera ture d u r i n g M R i m a g i n g are s p a t i a l l y var i ed a n d have suggested methods for r educ ing the effect o n M R I der ived dose m a p s [34]. However a reduced sens i t iv i ty to gel t emperature remains a n advantage of a n a l ternat ive read-out technique , x - r a y C T [39]. G e l t emperature has also been s h o w n to increase d u r i n g the p o l y m e r i z a t i o n process [90]. B e a m energy a n d dose rate have been shown to have no effect o n P A G dose response [91-93], a l t h o u g h energy does have a n effect o n other types of p o l y m e r gels [94]. However , par t i c l e type , i n p a r t i c u l a r p r o t o n a n d other heavy par t i c l e beams, can have a d r a m a t i c effect o n P A G M R I a n d C T dose responses [95, 96]. T h i s is exp la ined b y a decreased re lat ive effectiveness of the gel for h i g h L E T rad ia t i ons [97]. T e m p o r a l a n d S p a t i a l S t a b i l i t y Changes i n P A G M R I dose response over t ime were n o t e d a n d invest igated re la t ive ly ear ly o n b y several authors [98-100]. D e t a i l e d recent invest igat ions show two types of t e m p o r a l i n s t a b i l i t y : short a n d l o n g t e r m [101-103]. I n the short t e r m the slope of the dose response increases as p o l y m e r i z a t i o n proceeds. T h i s occurs u p to ~ 12 hours post i r r a d i a t i o n a n d p o l y m e r i z a t i o n is considered complete b y th is t i m e [101, 102]. T h e l ong p o l y m e r i z a t i o n p e r i o d is a t t r i b u t e d to l ong l i v e d macrorad i ca l s present i n the sys tem [98, 99]. A f t e r p o l y m e r i z a t i o n is complete , the sens i t iv i ty of the dose response remains stable , however a l o n g t e r m effect is observed i n the increase of the intercept or basel ine R 2 value over a p e r i o d of several months . T h i s is be l ieved to be due to ongo ing ge la t i on [101, 102]. S p a t i a l i n s t a b i l i t y of P A G dose d i s t r i b u t i o n s i n the f o r m of edge enhance-ments have been observed b y several authors [66, 68, 104, 105]. Invest igat ions b y D e D e e n e et al. demonstra ted a dependence of the effect o n p o s t - i r r a d i a t i o n t i m e [103]. T h i s effect has t y p i c a l l y been a t t r i b u t e d to di f fusion of monomers f r o m low to h i g h dose regions where t h e y react w i t h l ong l i v e d macrorad i ca l s [106]. M o d e l i n g w o r k has recent ly been used to es tab l i sh the l i n k between the observed t e m p o r a l a n d s p a t i a l ins tab i l i t i es a n d to show t h a t b o t h these effects c a n indeed be e x p l a i n e d b y the di f fusion hypothes is [88, 107, 108]. N e w T y p e s o f P o l y m e r G e l T h e development of new types of p o l y m e r gel dosimeters has recent ly become a n increas ing ly researched top ic . T h e use of different monomers is b e i n g invest igated i n hopes of i m p r o v i n g gel character is t i cs a n d r e d u c i n g gel tox i c i ty . F o r example , a new A^-v iny l pyrro l idone argon ( V i p a r ) gel was i n t r o d u c e d i n 1999 [109] w i t h the favourable character is t i c of a wide d y n a m i c dose range [110]. Several different monomers have been invest igated a n d Lepage et al. prov ide a n excel lent c o m p a r i s o n of r e s u l t i n g M R I dose responses [111]. T h e most sensit ive m o n o m e r i n th is s t u d y was m e t h a c r y l i c a c id , w h i c h appears a g a i n i n the first n o r m o x i c gel, as discussed below. A n o t h e r p r o m i s i n g gel is the h y d r o x y e t h y l a c ry la te ( H E A ) gel , recent ly invest igated i n d e t a i l b y G u s t a v s s o n et al. [112]. A n exc i t ing new area of research is the development of normoxic gels w h i c h can be m a n u f a c t u r e d i n the presence of oxygen. O x y g e n scavengers are used to chemica l ly de-oxygenate the gels, r e m o v i n g the need for a glovebox n i t rogen e n v i -ronment a n d great ly decreasing the c o m p l e x i t y a n d costs assoc iated w i t h gel m a n -ufacture . T h e first such gel , i n t r o d u c e d b y F o n g et al. , used the m e t h a c r y l i c a c i d m o n o m e r a long w i t h ascorbic ac id as oxygen scavenger a n d was t e r m e d the M A G I C gel [84]. T h i s work was fol lowed b y a de ta i l ed s t u d y of the f u n d a m e n t a l propert ies of several normox i c gels, us ing b o t h m e t h a c r y l i c a c i d a n d a c r y l a m i d e monomers , by DeDeene et al. [113]. Since th is t i m e several authors have invest igated n o r m o x i c gel f o rmulat ions , i n c l u d i n g use of a l ternate ge l l ing agents a n d oxygen scavengers [114-118]. 3.3.2 G e l I m a g i n g M o d a l i t i e s G e l i m a g i n g is another h i g h l y act ive area of gel d o s i m e t r y research. D u e largely to the expense a n d d i f f i cul ty accessing M R I scanners, m u c h w o r k has been done i n recent years i n f ind ing a l ternat ives to M R I i m a g i n g for p o l y m e r gels. I n p a r a l l e l , w o r k has cont inued i n the improvement of M R I techniques for p o l y m e r gel dos imetry . Deve lopment of M R I for p o l y m e r gel d o s i m e t r y is focussed large ly o n M R I techniques, charac ter i za t i on of M R I dose responses (as discussed i n the previous section) a n d so lut ions to i m a g i n g prob lems , such as artefacts . O p t i m i z i n g M R I se-quences for gel dos imetry has been shown to be c r i t i c a l for i m p r o v e d dose reso lut i on a n d i m a g i n g speed [119-121]. 3 D M R i m a g i n g techniques have also been a t t e m p t e d [122]. Several studies have addressed image noise [123, 124] a n d image d i s t o r t i o n [125] i n M R I p o l y m e r gel dosimetry . However , the most r igorous d iscuss ion of ar te -facts affecting gel dos imetry measurements b y M R I is f ound i n a series of excellent papers by DeDeene et al. [32-34]. These works prov ide strategies for r e d u c i n g M R I artefacts a n d recent a p p l i c a t i o n papers b y th i s group (h igh l ighted i n the fo l low-i n g section) show t h a t excellent M R I gel d o s i m e t r y results can be o b t a i n e d w h e n extreme care is taken to account for these p o t e n t i a l di f f icult ies [74, 105, 126]. A technique for h igh - reso lut i on m i c r o - M R I i m a g i n g has also been invest igated [127]. M a n y p o l y m e r gel systems undergo a n o p t i c a l change u p o n i r r a d i a t i o n , as was shown i n figure 1.12 i n chapter 1. T h i s character i s t i c is exp lo i t ed i n develop-ment of o p t i c a l s cann ing for dose read-out i n p o l y m e r gel dos imetry . T h i s technique was first i n t r o d u c e d for p o l y m e r gels i n two c o m p a n i o n papers b y G o r e et al. a n d M a r y a n s k i et al. i n 1996 [36, 37]. Since th i s t i m e , o p t i c a l C T ( O C T ) has emerged as a v iab le a l ternat ive to M R I for gel dos imetry , as s h o w n i n a c o m p a r i s o n made b y O l d h a m et al. [14]. T h e first O C T scanners, o p e r a t i n g i n a s i m i l a r fashion to first generat ion C T scanners (see sect ion 2.1.2), were t y p i c a l l y very slow, t a k i n g ~ 20 m i n to image a single slice [14]. However , a recent, c o m m e r c i a l l y avai lable ( M G S Inc.) O C T scanner demonstrates s l i gh t ly i m p r o v e d per formance character is -t ics [128, 129]. I n a d d i t i o n , several groups are deve lop ing O C T scanners t h a t use 2 D C C D camera detect ion systems i n order to improve s cann ing speed [130, 131]. Recent works b y O l d h a m et al. i n c l u d e basic invest igat ions charac ter i z ing the per -formance of O C T gel dos imetry a n d h igh l ight t h a t m u c h work , p a r t i c u l a r l y i n the e l i m i n a t i o n of artefacts , is s t i l l r e q u i r e d [132, 133]. I n a d d i t i o n , at th is t i m e O C T gel dos imetry is l i m i t e d to u n i f o r m c y l i n d r i c a l p h a n t o m s due to c o m p l e x scat ter ing pat terns i n t r o d u c e d b y n o n - s y m m e t r i c a l a n d / or heterogeneous p h a n t o m s . T h e re-cent gel dos imetry conference, D O S G E L 2 0 0 4 , h a d several papers o n O C T p o l y m e r gel dos imetry [134, 135], i n c l u d i n g o p t i c a l s cann ing of a n ent i re ly new p o l y u r e t h a n e dosimeter [136], c on f i rming cont inued interest i n this read-out technique. It shou ld be no ted that O C T is also a h i g h l y successful means of i m a g i n g r a d i o c h r o m i c , Fr i cke based gel systems [137]. O t h e r techniques for i m a g i n g p o l y m e r gel have also been recent ly in t roduced . P e r h a p s the most p r o m i s i n g , due to m a n y p r a c t i c a l advantages , is x - r a y C T . Since shown i n 2000 to be a feasible m e t h o d of r e a d i n g out gel dosimeters [39] interest i n C T gel dos imetry has been s tead i ly increas ing . Several groups have now t a k e n u p research i n this area, p r o d u c i n g qui te a n u m b e r of p u b l i s h e d papers a n d conference contr ibut i ons [41, 43, 96, 116, 118, 138-142]. T h e fo l l owing sect ion, 3.4, is solely devoted to C T gel dos imetry , the focus of th is thesis , a n d these recent works are discussed i n some de ta i l there. F u r t h e r m o r e , the work presented i n th is thesis has p r o d u c e d three research papers (two p u b l i s h e d a n d one submi t ted ) as w e l l two conference presentat ions [143-147]. A group i n A u s t r a l i a has shown t h a t u l t r a s o u n d (US) is another p o t e n t i a l read-out technique for p o l y m e r gel d o s i m e t r y [40, 148] due to changes i n the U S a t t e n u a t i o n coefficients w i t h dose i n P A G dosimeters [149]. I n the i r most recent work , they have shown i n t i a l success at i m a g i n g a s i m p l e dose d i s t r i b u t i o n [150]. F i n a l l y , R a m a n spectroscopy, i m p o r t a n t i n f u n d a m e n t a l p o l y m e r gel dose response studies [38, 80, 86, 97, 151] has recent ly s h o w n p o t e n t i a l for h i g h reso lut ion gel d o s i m e t r y [70]. 3.3.3 G e l A p p l i c a t i o n s A p p l i c a t i o n to the 3 D dose ver i f i ca t i on of r a d i a t i o n t h e r a p y t reatments is c l ear ly the f u n d a m e n t a l goal of p o l y m e r gel d o s i m e t r y a n d researchers have been inves t iga t ing such app l i ca t i ons since the ear ly days [67-69]. Since th is t i m e , p o l y m e r gels have been app l i ed , w i t h v a r y i n g degrees of success, to a range of r a d i o t h e r a p y a p p l i c a -t ions . Severa l authors have p r o v i d e d reviews of gel d o s i m e t r y app l i ca t i ons i n the l i t e ra ture [91, 152, 153] however these art ic les are now somewhat ou tdated . T h e fo l l owing provides a n u p d a t e d rev iew of l i t e ra ture o n the a p p l i c a t i o n of p o l y m e r gel dosimeters . It s h o u l d be n o t e d t h a t , w i t h few exceptions [126, 154], these works focus on v a l i d a t i n g the dosimeter per formance for a g iven a p p l i c a t i o n ra ther t h a n v a l i d a t i n g the t reatment itself. A l t h o u g h p o l y m e r gel dos imetry has been app l i ed to re la t ive ly s t a n d a r d t r e a -ments [154], app l i ca t i ons to c o m p l e x e x t e r n a l b e a m t reatments are most c o m m o n . I M R T is a p r i m e target for advanced dos imetry techniques a n d m u l t i p l e authors have a p p l i e d p o l y m e r gel to th is t r e a t m e n t moda l i ty . A n ear ly assessment of the p o t e n t i a l for I M R T gel dos imetry was g iven b y L o w et al. [155], but a more recent a n d comprehensive look at the requirements of I M R T ver i f i ca t i on a n d the a b i l i t y of p o l y m e r gel systems to meet these requirements is g iven b y DeDeene et al. [156]. Since th is t i m e , several authors have a p p l i e d M R I gel d o s i m e t r y to I M R T dose ver i f i -c a t i o n a n d a v a r i a t i o n thereof, in tens i ty m o d u l a t e d arc t h e r a p y ( I M A T ) , w i t h overa l l excellent results [74, 105, 126, 157]. A n o t h e r h i g h prec i s ion t rea tment , stereotac-t i c radiosurgery, has also been the focus of a n u m b e r of a p p l i c a t i o n p u b l i c a t i o n s [73, 125, 158-163], i n c l u d i n g one u s i n g C T gel dos imetry [41]. B r a c h y t h e r a p y has also seen a n u m b e r of gel dos imetry app l i ca t i ons for b o t h source charac ter i za t i on a n d dose d i s t r i b u t i o n measurement . Several authors have i n -vest igated p o l y m e r gel d o s i m e t r y for h i g h dose rate ( H D R ) b r a c h y t h e r a p y [104, 1 6 4 -166], n o t i n g a reduced dosimeter response at short r a d i a l distances f r o m the source. I n a d d i t i o n , p o l y m e r gels have been eva luated for low dose rate ( L D R ) b r a c h y t h e r -apy [92], low p h o t o n energies relevant to b r a c h y t h e r a p y [72], as w e l l some specific b r a c h y t h e r a p y t reatment techniques [167-169]. P o l y m e r gels have also been a p p l i e d to p r o t o n a n d heavy i o n therapies [170, 171]. However , th i s remains a di f f icult prospect due to decreased re lat ive effectiveness of the dosimeter close to t racks of heavy part ic les [97]. O t h e r novel app l i ca t i ons inc lude b o r o n n e u t r o n capture ther -a p y (and the response of p o l y m e r gel t o e p i t h e r m a l neutrons) [172-174] a n d t h e r m a l t h e r a p y [175]. T h e large a n d var ied b o d y of l i t e ra ture p e r t a i n i n g to p o l y m e r gel dos ime-t r y app l i ca t i ons confirms the value a n d versa t i l i t y of the technique. I n a d d i t i o n , the recent gel dos imetry conference, D O S G E L 2 0 0 4 , h a d m a n y cont r ibut i ons o n the app l i ca t i ons of p o l y m e r gel d o s i m e t r y u s i n g M R I [176-183] a n d O C T [134, 135], i n c l u d i n g one novel a p p l i c a t i o n to d iagnost i c i m a g i n g [142]. T h e large n u m b e r of these very recent papers h igh l ights the increas ing relevance of such work . 3.4 C T Polymer Gel Dosimetry T h i s sect ion provides some de ta i l ed b a c k g r o u n d re lated to C T read-out of p o l y m e r gel dosimeters. T h e re la t ionship between gel dens i ty a n d contrast i n gel C T images is descr ibed i n sect ion 3.4.1. Sec t ion 3.4.2 reviews prev ious w o r k i n C T gel dos imetry . F i n a l l y , a s u m m a r y of the advantages a n d l i m i t a t i o n s of us ing C T for gel read-out is p r o v i d e d i n sect ion 3.4.3. 3.4.1 D e n s i t y a n d C T N u m b e r A s descr ibed i n sect ion 2.1.3, the in tens i ty of C T images is expressed as a C T number (NCT > H ) w h i c h is a ra t i o of m a t e r i a l a t t e n u a t i o n coefficient re lat ive to water , as g iven b y equat i on 2.4. C o n t r a s t is observed i n C T images of i r r a d i a t e d p o l y m e r gel since a change i n NCT {ANCT ) occurs between i r r a d i a t e d a n d n o n - i r r a d i a t e d gel. T h i s change i n image in tens i ty is due to a change i n the l inear a t t e n u a t i o n coefficient of the gel , Apgei, w i t h i r r a d i a t i o n . F r o m the de f in i t i on of NCT > equat i on 2.4, ANCT observed i n the C T image of a n i r r a d i a t e d gel c a n be expressed as A NCT = 1000 x ^ti^L. (3.21) Pw It is we l l k n o w n t h a t p, c a n be a p p r o x i m a t e d b y p, = Neaep, where NE is the n u m b e r of electrons per g r a m , ae is the e lectronic cross sect ion a n d p is the p h y s i c a l dens i ty [184]. A s discussed b y T r a p p et al [140], i n a p o l y m e r gel sys tem, p is the o n l y component of p t h a t is expected t o change w i t h i r r a d i a t i o n . A s a result , ANCT can be re la ted to i r r a d i a t e d a n d n o n - i r r a d i a t e d gel densit ies (p\ a n d po, respect ively) b y the expression A A T C T = ( J E 0 + 1 0 0 0 ) ( ^ - 1 ) (3.22) Po where HQ is the NCT for a n u n i r r a d i a t e d gel [140]. F r o m this expression (equat ion 3.22), ANCT c a n D e d i r e c t l y re la ted to the density change o c c u r r i n g i n the gel u p o n i r r a d i a t i o n (Apgei) by ANCT = ( ^ ^ ) A P G E , (3.23) Po T h a t is , Apgei is g iven by A P G E L = KANCT (3.24) where, for a g iven system, K is a constant . B a s e d on d a t a r e p o r t e d i n the l i t e r a t u r e , for a P A G system K ~ 0.995 H k g - 1 m 3 [42, 71, 140]. T h i s means that Apgel, w h e n expressed i n u n i t s of k g m - 3 , is a p p r o x i m a t e l y n u m e r i c a l l y equa l to ANCT i n u n i t s of H . T h i s conversion of ANCT to Apgei is used i n the P A G c o m p o s i t i o n studies i n chapter 5. 3.4.2 L i t e r a t u r e R e v i e w T h i s sect ion provides a de ta i l ed rev iew of l i t e ra ture o n C T gel dos imetry . T h e con-cept of us ing x - r a y C T for read-out of i r r a d i a t e d p o l y m e r gels was first i n t r o d u c e d b y H i l t s et al. at the first i n t e r n a t i o n a l gel dos imetry conference, D O S G E L 9 9 , a n d i n a subsequent paper [39, 185]. I n a d d i t i o n , th is w o r k compr i sed M . H i l t s ' M S c thesis [42]. These works showed the first C T dose response for a P A G dosimeter , h i g h l i g h t e d a low t e m p e r a t u r e dependence of the sys tem a n d discovered the i m p a c t of x - r a y tube heat ing o n the dose response. These earl ier studies also d e m o n s t r a t e d t h a t image averaging a n d b a c k g r o u n d s u b t r a c t i o n are i m p o r t a n t tools for p r o d u c i n g q u a l i t y dose maps . F o l l o w - u p w o r k showed the a b i l i t y of C T gel dos imetry to cor-rec t ly local ize the h i g h dose reg ion de l ivered b y a stereotact ic rad iosurgery t rea tment a n d the h i g h spa t ia l r eso lut i on capabi l i t i es of the technique t h r o u g h measurement of a c l in i ca l p r o t o n b e a m P D D [41, 96]. W i t h the proven feas ib i l i ty of C T gel dos imetry , a group i n A u s t r a l i a (headed b y D r . C l i v e B a l d o c k ) , became ac t ive ly invo lved i n C T gel research. T h e i r first p u b l i c a t i o n invest igated the dose response for different f o rmulat i ons of P A G [43]. I n p a r t i c u l a r , th i s w o r k showed a n increase i n the C T dose response w i t h P A G % T , b u t no change i n sens i t iv i ty w i t h ge la t in concentrat ion . T h e y d i d not however invest igate the effect of P A G % C o n the dose response. I n a fo l low-up work , th is group made direct measurements , independent of C T scanning , to invest igate the change i n P A G l inear a t t enuat i on coefficient w i t h dose. T h e y also made e x p e r i m e n t a l measures of the change i n P A G dens i ty w i t h dose. These i m p o r t a n t exper iments proved the hypothes is tha t C T contrast i n i r r a d i a t e d P A G is due to a s m a l l dens i ty change t h a t occurs i n these systems [140]. I n a d d i t i o n a l works presented b y th i s group at D O S G E L 2 0 0 1 , h igh l ights inc lude a r o u g h c a l c u l a t i o n showing t h a t four t imes the cur rent ly observed P A G dens i ty change is a l lowable before s p a t i a l d i s tor t ions prevent the system f rom meet ing the I n t e r n a t i o n a l C o m m i s s i o n o n R a d i a t i o n U n i t s a n d Measurements ( I C R U ) 0.2 c m s p a t i a l r eso lut i on requirement [186] as w e l l as a n analys is of factors affecting dose u n c e r t a i n t y i n C T gel d o s i m e t r y [138, 139]. A p a r t f r om the pub l i ca t i ons a n d conference presentat ions based on the w o r k presented i n this thesis [143-147] the most recent l i t e ra ture o n C T gel d o s i m e t r y are several works presented at th is year 's gel d o s i m e t r y meet ing , D O S G E L 2 0 0 4 [116, 141, 142], a n d a p u b l i c a t i o n shor t l y thereafter [118]. T w o of these works inves-t i ga ted the response of n o r m o x i c P A G a n d M A G I C gel dosimeters to C T read-out [116, 118]. T h i s is i m p o r t a n t w o r k as i t paves the w a y for a v e r y p r a c t i c a l m e t h o d of 3 D dose measurement a n d therefore perhaps introduces the next m a j o r s t r ide f o rward i n gel dos imetry : C T read-out c o m b i n e d w i t h bench- top gel manufac ture . However , these i n i t i a l studies showed conf l i c t ing results : one d e m o n s t r a t e d the re -sponse of normox i c P A G to be near ly i d e n t i c a l t o s t a n d a r d P A G [116] a n d the other showed a signif icant r educ t i on i n response [118]. T h e most probab le reason for th i s d iscrepancy is different amounts of oxygen scavenger used i n gel p r e p a r a t i o n . T h i s h ighl ights the i m p o r t a n c e of o p t i m i z i n g gel f o rmulat i ons . A n o t h e r w o r k focused o n C T hardware for gel dos imetry u s i n g a benchtop , first generat ion t y p e C T sys-t e m [141]. T h i s w o r k w i l l hope fu l ly l ead to va luab le h a r d w a r e recommendat ions for o p t i m i z e d C T gel dos imetry , however, use of such a sys tem removes the largest advantage of C T read-out , the access ib i l i ty a n d speed of c o m m e r c i a l C T systems. A f ina l recent work invest igated the use of M A G I C gel t o measure the dose d i s t r i b u t i o n produced b y C T scanners [142]. T h i s w o r k is c o m p l e m e n t a r y to invest igat ions of C T read-out for gel dos imetry since C T dose has the p o t e n t i a l to affect i r r a d i a t e d gels i f they are not inert w h e n imaged . 3.4.3 A d v a n t a g e s a n d L i m i t a t i o n s T h e r e are m a n y advantages to u s i n g C T ins tead of M R I or O C T for gel dos ime-try . M a n y of these are p r a c t i c a l advantages for c l i n i c a l i m p l e m e n t a t i o n . P e r h a p s the most s igni f icant is the wide access ib i l i ty of C T scanners to r a d i a t i o n t h e r a p y departments . A s discussed i n chapter 1, C T s i m u l a t i o n is a n i m p o r t a n t component of m o d e r n r a d i a t i o n t h e r a p y pract i ce a n d as such m a n y cancer hospi ta ls have one or more C T scanners a l ready w i t h i n the ir r a d i a t i o n t h e r a p y departments . I n concert w i t h th is accessibi l i ty , C T scanners are fast a n d easy to operate . These facts m e a n that C T read-out c o u l d a l low m e d i c a l phys ic is ts i n m a n y cancer hospi ta ls to p e r f o r m gel dos imetry regular ly . I n terms of dos imeter qual i ty , C T read-out offers images w i t h s m a l l voxe l size, a l ow dependence of response on i m a g i n g t e m p e r a t u r e a n d a low level of artefacts w h i c h , thus far, c a n be successfully removed b y b a c k g r o u n d subt rac t i on . T h e r e are however some disadvantages. T h e most serious is a n i n a b i l i t y to detect s m a l l changes i n dose. A s discussed i n more d e t a i l i n the fo l l owing sect ion, 3.5, th i s is due to l ow dos imeter sens i t iv i ty a n d h i g h image noise a n d inves t iga t ing these issues is one focus of th is thesis . A n o t h e r p o t e n t i a l d isadvantage is t h a t , u n l i k e M R I or O C T read-out , C T scanners del iver dose to the gel so tha t the gel must be inert pr i o r t o i m a g i n g . T h u s far th i s has been accompl i shed b y re -oxygenat ing the gel post i r r a d i a t i o n , a technique t h a t m a y prove dif f icult for large phantoms . 3.5 Dose Resolution in C T Gel Dosimetry A s descr ibed i n sect ion 1.3.1, dose reso lut ion , or the m i n i m u m detectable change i n dose, is a n i m p o r t a n t character i s t i c of r a d i a t i o n t h e r a p y dosimeters . Since i m p r o v i n g the poor dose reso lut i on i n C T gel dos imetry is a c e n t r a l goal of th i s thesis , a further i n t r o d u c t i o n to the concept , w i t h p a r t i c u l a r emphasis o n C T gel dos imetry , is g iven here. Dose reso lut i on a n d u n c e r t a i n t y i n gel dos imetry has been discussed i n several pub l i ca t i ons [112, 119, 124, 187, 188] a n d th is d iscuss ion draws large ly f r o m these works . 3.5.1 A b s o l u t e D o s e R e s o l u t i o n Dose reso lut ion , des ignated as D^, represents the m i n i m u m dose difference (DA, G y ) t h a t can be d i s t ingu i shed w i t h a g iven level of confidence ( P ) [119]. I n general , D£ is re lated to the s t a n d a r d d e v i a t i o n i n a dose measurement ((TD) by: D% = kpV2aD (3.25) where kp is the coverage factor (parameter of a S tudent ' s T d i s t r i b u t i o n ) . F o r a 9 5 % confidence level (i.e. i n order to k n o w the value of w i t h 9 5 % confidence) , kp is 1.96 a n d - D A 5 % , f r o m equat i on 3.25, is g iven b y 2.77ao [189]. I f ory is used to a p p r o x i m a t e £>A, as is f requent ly done, the degree of confidence i n the dose d e t e r m i n a t i o n is ~ 6 7 % . I n C T gel dos imetry , dose is de te rmined f r o m a measurement of NCT t h r o u g h the a p p l i c a t i o n of a calibration curve tha t relates dose to NCT- T h i s c a l i b r a t i o n curve is der ived f rom a dose response curve w h i c h is the result of measurements of NCT as a func t i on of dose. I n general , for a dose (D) g iven b y D = f(NCT) (3.26) b y s t a n d a r d T a y l o r expans ion , dose uncerta inty , ao, is g iven b y '•> - f " " 2 0 2 + " 2 ( § > 2 <*•"> where O~NCT * s * n e u n c e r t a i n t y i n the measurement of NCT a n d i a n d CTJ are the fit parameters a n d the i r assoc iated uncerta int ies . Severa l authors have shown t h a t gel sens i t iv i ty a n d u n c e r t a i n t y i n the measurement (for C T gel dos imetry , sfj^ a n d < T A T c t , respectively) are the most s ignif icant factors i n d e t e r m i n i n g dose reso lut i on [119, 190]. Frequent ly , a l inear dose response (or a quas i - l inear reg ion of a dose response) is used i n gel dosimetry . F o r these l inear cases, ary is r e la ted to G^CT by the slope (or sens i t iv i ty ) of the response, sfJ^T • I n a d d i t i o n , t h e change i n i m a g i n g parameter (e.g. ANCT) is often measured so tha t the intercept is zero. I n these cases, a c a l c u l a t i o n of cro reduces to: A D = [ 6 N ^ T ] < 7 ^ ( 3 - 2 8 ) where NCT 1S measured above the background , u n i r r a d i a t e d value. C o m b i n i n g equa-t ions 3.25 a n d 3.28 al lows for c a l c u l a t i o n of abso lute dose reso lut ion , D ^ , i n G y . 3.5.2 R e l a t i v e D o s e R e s o l u t i o n O f t e n i n gel dos imetry i t is a relative or percentage dose reso lut ion , Z?£ % w h i c h is of interest . % is g iven b y Di,% = 100 x (3.29) where D N O R M is the n o r m a l i z a t i o n dose. T h e n o r m a l i z a t i o n dose is the m a x i m u m i r r a d i a t e d dose, i n G y , a n d is equal to 100%, re la t ive dose. T h e best dose reso lut ion DP over the dose range (0 G y to DNORM) is achieved b y choos ing D N O R M to m i n i m i z e -fi-[112]. C o m b i n i n g equations 3.25, 3.28 a n d 3.29 we see t h a t re lat ive dose reso lut ion , £>£ %> ^ i m p r o v e d b y (1) increas ing the response of NCT to dose (i.e. decreasing ), (2) increas ing the dose range (i.e. DNORM) a n d (3) reduc ing <JNCT- T h i s thesis investigates a n d furthers our u n d e r s t a n d i n g o f w h a t determines the na ture of the NCT to dose response a n d develops means of r e d u c i n g o-^CT. Chapter 4 Materials and Methods T h i s chapter describes mater ia l s a n d methods used t h r o u g h o u t th is thesis. P r o -cedures a n d equipment used for gel manufac ture , gel i r r a d i a t i o n , C T i m a g i n g a n d image processing a n d d a t a analys is are deta i led i n sect ions 4.1, 4.2, 4.3 a n d 4.4, respectively. A d d i t i o n a l detai ls of p h a n t o m s , C T i m a g i n g a n d analys is techniques specific t o the i n d i v i d u a l studies i n th i s thesis are p r o v i d e d at the b e g i n n i n g of chapters 5 t h r o u g h 7. 4.1 Gel Manufacture A s was descr ibed i n chapter 3, the response of p o l y m e r gel t o r a d i a t i o n is i n h i b i t e d b y oxygen a n d P A G must be m a n u f a c t u r e d i n a n anox i c e n v i r o n m e n t . I n th is w o r k a l l gels were m a n u f a c t u r e d i n a n i t rogen glovebox a n d a d isso lved oxygen meter was used to m o n i t o r oxygen levels i n b o t h the box a n d the gel . T h e glovebox a n d oxygen meter are de ta i l ed i n sections 4.1.1 a n d 4.1.2. Sec t i on 4.1.3 describes the gel m a n u f a c t u r i n g procedure a n d p h a n t o m s used to house the gels. 4.1.1 N i t r o g e n A t m o s p h e r e G l o v e B o x T h e n i t rogen atmosphere glovebox used i n these exper iments was designed b y a previous student i n our lab [191] a n d b u i l t in -house . T h e b o x houses a p p r o x i m a t e l y 25 L of gas a n d is constructed of 0.8 c m clear Perspex . O n e end of the b o x has a large rectangular open ing for p l a c i n g suppl ies ins ide . T h i s open ing is sealed b y a door w h e n the glovebox is i n use. A set of gas i m p e r m e a b l e gloves are a t t a c h e d to two 20 c m d iameter holes on the front of the b o x i n order to a l low access d u r i n g gel p r e p a r a t i o n . G a s is supp l i ed f r om a n i t r o g e n t a n k v i a a c r y l i c t u b i n g a n d pressure is contro l l ed by a c o m b i n a t i o n of t a n k regulators a n d a flow meter m o u n t e d o n the glovebox. T h e gas out let is a one w a y valve t h a t is set to create a sl ight pos i t ive pressure w i t h i n the box , a n d hence a v o i d oxygen f r o m re-enter ing v i a the out let t u b i n g . E q u i p m e n t p e r m a n e n t l y ins ide the b o x inc ludes a fan , a n oxygen meter , descr ibed below, sect ion 4.1.2, a n d a n e lectronic heater a n d s t i r p late . 4.1.2 O x y g e n M e t e r O x y g e n levels w i t h i n the glovebox a n d the l i q u i d gel were measured u s i n g a n Oxi® 340 dissolved oxygen meter ( W T W Wissenschaf t l i ch -Technische W e r k s t a t t e n G m b H , W e i l h e i m , G e r m a n y ) . A schematic d i a g r a m of the meter is shown i n figure 4.1. T h e meter consists of a s m a l l , 17 x 8 c m h a n d h e l d device w i t h a d i g i t a l d isplay , a Cel lOx® 325 waterproo f probe ~ 15 c m l o n g a t t a c h e d to the meter b y a 1.5 m l o n g cable. T h e probe head contains the oxygen sensor. T h e sensor consists of two elec-trodes submerged i n a n electrolyte so lu t i on . A gas permeable membrane separates the so lu t i on a n d electrodes f r om the sample . O n e electrode reduces oxygen molecules to hydrox ide ions a n d i n th is e lec trochemica l r eac t i on a current flows between the electrodes. T h e current depends o n the concentra t i on of oxygen [O2] i n the sample a n d the meter calculates [O2] f r om the measured current . T h e measurab le range of [0 2 ] is f r o m 0 to 19.99 mg /1 w i t h a r eso lu t i on of 0.5 % [192]. T h e meter also measures t emperature w i t h i n the range 2 to ~ 40°. I n gel manufac ture the meter is waterproof probe cable probe head electrodes electrolyte \ solution oxygen permeable membrane digital meter F i g u r e 4.1: A schematic d i a g r a m showing the m a i n components of a dissolved oxygen meter . used to measure [O2] i n the glovebox env i ronment a n d [O2] a n d t e m p e r a t u r e w i t h i n the l i q u i d gel. Before use the meter is c a l i b r a t e d i n a i r u s i n g a W T W air c a l i b r a t i o n vessel designed speci f ical ly for th is purpose . 4.1.3 G e l M a n u f a c t u r e T e c h n i q u e A consistent procedure was used to m a n u f a c t u r e the gels for a l l exper iments . Ge l s were composed of electrophoresis grade a c r y l a m i d e monomer a n d N, A^'methylene b i s a c ry l a m i de (bis) crossl inker ( S i g m a - A l d r i c h , St . L o u i s M O , U S A ) , ge la t in (300 B l o o m , S i g m a - A l d r i c h ) a n d de ion ized water a n d were prepared w i t h i n a n in-house designed a n d b u i l t glovebox t h a t is descr ibed above, sect ion 4.1.1. R e q u i r e d amounts of d r y chemicals were measured in to s m a l l glass flasks u s i n g a prec is ion scale. T h e de ionized water was p laced in to a 500 m L glass flask t h a t was used to house the gel d u r i n g p r o d u c t i o n . A l l mater ia l s r equ i red for gel manufac ture were t h e n sealed w i t h i n the glovebox a n d the sys tem was p u r g e d w i t h d r y n i t rogen ( N 2 ) t o a achieve [O2] r ead ing less t h a n 0.05 m g / L u n i f o r m l y t h r o u g h o u t . [O2] was measured u s i n g a n O x i 340 dissolved oxygen meter t h a t is de ta i l ed above, sect ion 4.1.2. A f t e r glovebox p u r g i n g (~60 m i n ) , N 2 was b u b b l e d d i r e c t l y in to the water u n t i l the [O2] i n the water was less t h a n 0.01 m g / L . C o n t i n u o u s s t i r r i n g throughout ensured u n i f o r m de-oxygenat ion . T h e purged water was heated to 30° C a n d the ge la t in a n d a c r y l a m i d e (if required) were added . A t th i s t i m e direct N 2 b u b b l i n g was s t opped to prevent f r o t h i n g of the gel a n d N 2 p u r g i n g was cont inued by vapour exchange alone. T h e gel was further heated to 45°C a n d , after coo l ing to 43-44°C, the bis (if required) was added. H i g h e r temperatures are avo ided to prevent bis f r o m spontaneous ly p o l y m e r i z i n g . F i n a l l y , s t i r r i n g a n d N 2 p u r g i n g was cont inued u n t i l the bis was f u l l y d isso lved a n d the gel m i x t u r e c lear a n d u n i f o r m . W i t h i n the N 2 g lovebox env i ronment , the comple ted gel was t rans ferred into 10, 20 m L h i g h densi ty po lyethylene ( H D P E ) l i q u i d s c i n t i l l a t i o n v ia ls ( W h e a t o n Science P r o d u c t s , M i l l v i l l e N J , U S A ) . T h e v ia l s were i n d i v i d u a l l y sealed i n c y l i n -F i g u r e 4.2: S c i n t i l l a t i o n v ia ls a n d in-house made p h a n t o m s used i n p r e p a r a t i o n , storage a n d i r r a d i a t i o n of p o l y m e r gels. d r i c a l a c r y l i c phantoms (designed a n d b u i l t in -house) , removed f r o m the glovebox a n d set to gel i n the refr igerator for about 3 hours . T h e v ia l s a n d p h a n t o m s are p i c -t u r e d i n f igure 4.2. T h e r e were several considerat ions w h e n des igning this system. P l a s t i c v ia l s were chosen over the more c o m m o n glass s c i n t i l l a t i o n v ia ls i n order to reduce artefacts w h e n C T scanning . However , these t h i n w a l l e d (1 m m ) p last i c v ia ls are permeable to oxygen a n d i f left to gel i n the refr igerator the gels w o u l d not react w h e n i r r a d i a t e d . T h i s p r o b l e m was solved b y seal ing the v ia l s w i t h i n 2.54 c m t h i c k w a l l e d a c r y l i c p h a n t o m s , as p i c t u r e d i n f igure 4.2. T h e thickness of these p h a n t o m s was speci f ical ly designed to prov ide a n O2 bar r i e r d u r i n g i r r a d i a t i o n a n d p o s t - i r r a d i a t i o n p o l y m e r i z a t i o n based on oxygen p e r m e a b i l i t y d a t a i n the l i t e ra ture [193]. F u r t h e r m o r e , these acry l i c p h a n t o m s f o rm p a r t of the i r r a d i a t i o n p h a n t o m required to del iver u n i f o r m doses to the gel v ia l s , as descr ibed i n sect ion 4.2.2. U s e of the acry l i c phantoms meant that the v i a l wal ls themselves d i d not require O2 barr ier propert ies . I n fact, the oppos i te became desirable a n d H D P E was selected for i ts h i g h O2 p e r m e a b i l i t y [194]. T h i s was to encourage r a p i d di f fusion of O2 into the gels p o s t - p o l y m e r i z a t i o n i n order to render the gels inact ive for C T i m a g i n g . 4.2 Gel Irradiation A l l gels were i r r a d i a t e d u s i n g a m e d i c a l l inear accelerator or linac, the most c o m m o n device used for dose de l ivery i n r a d i a t i o n therapy. L i n a c s c a n produce p h o t o n a n d e lectron beams of energies r a n g i n g f r o m 4 to ~ 25 M e V . T h e basic components a n d operat i on of a l inac are descr ibed i n sect ion 4.2.1 a n d detai ls of the technique used to i r r a d i a t e gel v ia l s are p r o v i d e d i n sect ion 4.2.2. D e t a i l e d theory of l inac design a n d opera t i on m a y be found i n a t ex t b y C . J . K a r z m a r k [195]. A v e r y br ie f overview of l i n a c design is in c luded here as b a c k g r o u n d since i t does not have a direct i m p a c t on gel dos imetry . 4.2.1 M e d i c a l L i n e a r A c c e l e r a t o r A schematic d i a g r a m of a t y p i c a l l i n a c is s h o w n i n figure 4.3. A n e lectron g u n del ivers a pulsed b e a m of ~ 25 k e V electrons into a n acce lerat ing waveguide. These pulses are co inc ident w i t h ~ 3000 M H z microwaves p r o d u c e d b y a m a g n e t r o n a n d the electrons are accelerated t h r o u g h the waveguide . E l e c t r o n s ex i t the waveguide at energies of u p to 25 M e V a n d b e n d i n g magnets are used to steer a n d focus the b e a m . A v a c u u m is m a i n t a i n e d throughout the e lec tron t r a n s p o r t s y s t e m (electron gun , acce lerat ing waveguide a n d b e n d i n g magnet assembly) . T h e h i g h energy electrons enter w h a t is ca l led the t reatment head. T h e t r e a t m e n t head as s h o w n i n figure 4.3 is conf igured for a p h o t o n b e a m . P h o t o n s are p r o d u c e d b y B r e m s t r a h l u n g processes i n a target , t y p i c a l l y made of tungsten . T h e r a d i a t i o n b e a m passes t h r o u g h p r i m a r y co l l imators a n d t h e n a f la t ten ing f i lter w h i c h spreads out the f o r w a r d peaked p h o t o n b e a m to produce a u n i f o r m profi le . I on chambers m o n i t o r b e a m o u t p u t a n d s y m m e t r y . Secondary co l l imators define t h e f ield size covered b y the r a d i a t i o n . T h i s can be fo l lowed b y a m u l t i l e a f c o l l i m a t o r w h i c h , as descr ibed i n chapter 1, can be used to further shape r a d i a t i o n fields or to del iver c o m p l e x I M R T treatments . F o r a n e lectron b e a m the con f igurat ion o f the t r ea tment head is s i m i l a r w i t h the except i on that the target is replaced w i t h a s ca t ter ing f o i l to spread out the concentrated b e a m vacuum pumping system electron e" gun w pulsed modulator accelerating waveguide circulator RF waveguide microwave source (magnetron or klystron) electron transport (bending magnet) treatment head photon target primary collimators J flattening filter ion chambers secondary collimators tertiary collimators (e.g. multi-leaf collimator) F i g u r e 4.3: A schematic d i a g r a m of a m e d i c a l l inear accelerator ( l inac) . A s shown, the conf igurat ion i n the l inac head produces a p h o t o n b e a m . F o r e lectron b e a m p r o d u c t i o n the target is replaced w i t h a s cat ter ing fo i l . of electrons [6, 12, 195]. I n a d d i t i o n , the c o l l i m a t o r is supp lemented b y a n e lectron cone or app l i ca to r t h a t serves to define the b e a m edges a n d compensate for m u l t i p l e e lectron scat ter ing i n a i r between source a n d pat ient . 4.2.2 I r r a d i a t i o n T e c h n i q u e A l l gel samples i n th is thesis were i r r a d i a t e d u s i n g 6 M V p h o t o n beams f r o m C L 2 1 0 0 a n d C L 2 1 E X l inacs ( V a r i a n M e d i c a l Sys tems Inc . , P a l o A l t o C A , U S A ) . E a c h v i a l was i r r a d i a t e d i n d i v i d u a l l y , sealed i n i t s a c r y l i c p h a n t o m (descr ibed i n sect ion 4.1.3), i n a purpose -bu i l t a c r y l i c i r r a d i a t i o n p h a n t o m , shown i n figure 4.4. T h i s p h a n t o m was designed i n order to fac i l i tate de l ivery of a n accurate u n i f o r m dose to each v i a l . T h e p h a n t o m is square, p r o v i d i n g flat surfaces for b e a m ent ry at 90° angles. E a c h side measures 20 c m across. A c a v i t y for i n s e r t i n g the gels w i t h i n the i r c y l i n d r i c a l containers is at the centre. T h e c a v i t y was prec i s i on m a c h i n e d to m a t c h the d i a m -eter of the cy l inders so t h a t no a i r gaps are present i n the sys tem. T h e p h a n t o m dimensions are designed to p o s i t i o n the gel v i a l at 10 c m d e p t h for a l l b e a m entry po ints . T h i s is desirable since l i n a c beams are c a l i b r a t e d a n d b e a m profiles are f lattest at th is d e p t h [12, 16]. I n a d d i t i o n , a 20 x 20 c m 2 p h a n t o m provides enough scatter m a t e r i a l for use of s t a n d a r d s ized fields (e.g. 10 x 10 c m 2 ) wh i l e s t i l l b e ing p r a c t i c a l for repeated use. L i n e s engraved o n the p h a n t o m m a r k e d the centre of the v ia l s a n d a l lowed for reproduc ib le set -up to the l i n a c isocentre. A pro to co l for i r r a d i a t i n g the v ia l s was designed w i t h two goals: t o per f o rm m u l t i p l e i r rad ia t i ons re la t ive ly q u i c k l y a n d to prov ide a u n i f o r m dose to each gel v i a l . T h e gel i r r a d i a t i o n p h a n t o m , i n c l u d i n g a v i a l filled w i t h inac t i ve gel , was C T scanned a n d i m p o r t e d in to a c o m m e r c i a l t r ea tment p l a n n i n g software package used at the Vancouver C a n c e r Centre ca l led C a d P l a n ™ v6.4.7 ( V a r i a n ) . T h e i r r a d i a t i o n goals were achieved u s i n g two p a r a l l e l opposed l a t e r a l beams w i t h field size 10 x 10 c m 2 . A source to axis d istance of 100 c m a n d d e p t h of 10 c m were used. A n inhomogene i ty correc t ion ( M o d i f i e d B a t h o , V a r i a n ) was a p p l i e d to account for the F i g u r e 4.4: T h e set-up used for a l l p o l y m e r gel i r rad ia t i ons i n this study. E a c h gel v i a l was i r r a d i a t e d i n d i v i d u a l l y w i t h i n its c y l i n d r i c a l acry l i c p h a n t o m , p i c t u r e d i n figure 4.2, us ing the square i r r a d i a t i o n p h a n t o m as shown. densi ty of the acry l i c p h a n t o m . F i g u r e 4.5 shows the excellent dose u n i f o r m i t y achieved i n a l l three planes ( ax ia l , sag i t ta l a n d coronal) w i t h i n the gel v i a l . A ca l cu la ted dose vo lume h i s t o g r a m ( D V H ) i n d i c a t e d that 9 9 % of the v i a l vo lume received 100% of the prescr ibed dose, w i t h the m i n i m u m a n d m a x i m u m doses at 99 .5% a n d 100.6%, respectively. U s i n g this technique, for each gel b a t c h nine v ia l s were i r r a d i a t e d to doses r a n g i n g f r om 2 to 20 G y a n d one v i a l was left u n - i r r a d i a t e d . T h e i r r a d i a t e d v ia ls were removed f r om the i r a c r y l i c p h a n t o m s about 15 hours p o s t - i r r a d i a t i o n as p o l y -m e r i z a t i o n is complete by th is t i m e [101, 102]. T h e t h i n w a l l e d H D P E v ia ls are h i g h l y permeable to oxygen [194] a n d given the i r s m a l l size (27 m m diameter ) oxy -gen diffuses t h r o u g h the gel samples w i t h i n ~ 1 2 hours [85]. T h i s process was used to render the gels inact ive a n d prepare t h e m for C T scanning . (a) (b) (c) F i g u r e 4.5: T h e treatment, p l a n designed for i r r a d i a t i o n of gel v ia ls . C T p l a n n i n g was used a n d the ca l cu la ted dose d i s t r i b u t i o n is shown here super imposed on images of the v i a l i r r a d i a t i o n p h a n t o m i n three planes: a x i a l (a), coronal (b) a n d sag i t ta l (c). In a l l v iews the 99 % isodose l ine (red) encompasses the gel v i a l vo lume. 4.3 C T Imaging A G E H i S p e e d CT/i scanner ( G E M e d i c a l Systems, M i l w a u k e e W I , U S A ) was used for a l l C T i m a g i n g i n th i s thesis . Sec t i on 4.3.1 describes this scanner. D e t a i l s of C T i m a g i n g techniques used throughout th is thesis are p r o v i d e d i n sect ion 4.3.2. F u r t h e r detai ls of C T i m a g i n g p h a n t o m s a n d protoco ls specific to i n d i v i d u a l studies are p r o v i d e d i n chapters 5, 6 a n d 7. 4.3.1 C T S c a n n e r : G E H i S p e e d C T / i T h e G E H i S p e e d C T / z is a 3 r d generat ion C T scanner. A s descr ibed i n chapter 2, a n x - r a y t u b e a n d detector a r ray rotate s imul taneous ly w i t h i n the g a n t r y to col lect the pro j e c t i on d a t a . T h e basics of the o p e r a t i o n a n d components of th is t y p e of scanner are as descr ibed b y figure 2.3. T h e H i S p e e d scanner used i n th i s thesis is equ ipped w i t h a P e r f o r m i x ™ x - r a y tube . T h e target has a t u n g s t e n - r h e n i u m focal t rack on a m o l y b d e n u m al loy substrate backed o n to a graphi te base. T h e anode angle is 7° a n d the t u b e operates w i t h a n o m i n a l focal spot size of e i ther 0.7 x 0.6 m m 2 or 0.9 x 0.9 m m 2 depend ing o n t u b e l oad . T h e x - r a y t u b e has the equivalent of 4.75 m m A l f i l t r a t i o n i n b o t h a d d e d t u b e f i l t r a t i o n a n d f ixed f i l t r a t i o n due to the detector co l l imators . T h e H i S p e e d detector is a n 864 channel a r ray of so l id state s c in t i l la tors coupled to photodiodes . A schemat ic d i a g r a m of the detector a r ray is shown i n figure 4.6. There are 852 act ive , image generat ing elements a n d 12 reference elements. T h e sc int i l la tors , ca l l ed HiLight® detectors, are made of a p o l y c r y s t a l l i n e ceramic m a t e r i a l . T h i s p r o p r i e t a r y m a t e r i a l is un ique a m o n g C T scanners a n d is r e p o r t e d to offer l ow afterglow, excellent z -axis un i f o rmi ty , h i g h transparency , g o o d x - r a y s topp ing a b i l i t y a n d h i g h l ight o u t p u t . These character ist i cs result i n stable , h i g h l y efficient detectors w i t h Q D E above 9 8 % [47]. E a c h s c in t i l l a tor is 1 x 20 x 20 m m 2 . F o r a l l scans the entire detector w i d t h is act ive a n d c o l l i m a t i o n of the x - r a y b e a m , focused on the detector centres, determines slice thickness . single element, 1 mm wide 20 mm L 852 detector elements 49° arc F i g u r e 4.6: Schemat ic d i a g r a m of the H i L i g h t detector a r r a y used i n the G E H i S p e e d C T / i C T scanner. T h e d a t a acqu is i t i on sys tem samples each detector at a rate of 984 H z , a m p l i -fies a n d quantif ies the current a n d t r a n s m i t s th i s i n f o r m a t i o n to the image generat ing system. T h e r e the image is reconstructed f r o m the p r o j e c t i o n v iews u s i n g a f i l tered b a c k p r o j e c t i o n a l g o r i t h m as descr ibed i n sect ion 2.3.3. Image recons t ruc t i on is fast a n d images are d isp layed on the console c o m p u t e r i n near rea l - t ime . W i n d o w a n d level ad justments at the console a l l ow for o p t i m i z a t i o n of v i ewed image contrast . 4.3.2 C T I m a g i n g T e c h n i q u e F o r a l l exper iments phantoms were p laced o n the C T couch, level led a n d centred w i t h i n the bore of the C T scanner. P h a n t o m s were s u p p o r t e d b y c u s t o m made s tyro foam j igs w h e n required . L a t e r a l a n d s a g i t t a l lasers were used for reproduc ib le pos i t i on ing . I n a l l cases, the i m a g i n g slice was set i n the m i d d l e of the l o n g i t u d i n a l axis of the p h a n t o m . Deta i l s of the specific C T p h a n t o m s used i n each s t u d y are p r o v i d e d i n chapters 5, 6 a n d 7. A f t e r r u n n i n g a s t a n d a r d p r e - p r o g r a m m e d set of w a r m - u p sequences a series of 100 images was scanned. These images were not r e ta ined but used solely to ensure a consistent scanner operat ing t e m p e r a t u r e across a l l C T i m a g i n g sessions. T u b e Tab le 4.1: I m a g i n g parameters avai lable o n the G E H i S p e e d CT/i C T scanner. D i a l u p scan parameters (units) Poss ib le values t empera ture is a n i m p o r t a n t factor i n ach iev ing reproduc ib l e C T dose responses i n gel dos imetry [39, 42]. S c a n protoco ls were selected f r o m the s canning parameters avai lable o n the G E H i S p e e d . These are l i s t ed i n tab le 4.1. T h e p a r t i c u l a r i m a g i n g protoco ls used var i ed w i t h each s t u d y a n d these detai ls are p r o v i d e d w i t h i n the respective chapters , 5, 6 a n d 7. F o r a l l image sets a second set of b a c k g r o u n d images was ob ta ined u s i n g the same p r o t o c o l i n order to p e r f o r m a b a c k g r o u n d subt rac t i on . 4.4 Image Processing and Data Analysis A l l C T images were t rans ferred over, a ne twork to a P C where they were u p l o a d e d v i a a n in-house d i c o m reader. A l l subsequent image process ing a n d analys is was per formed u s i n g software w r i t t e n i n M a t L a b ( T h e M a t h W o r k s Inc. , N a t i c k M A , U S A ) . Image averaging (if required) a n d b a c k g r o u n d s u b t r a c t i o n were per formed . B a c k g r o u n d s u b t r a c t i o n is c r i t i c a l for r e m o v i n g artefacts t h a t m a y d is tor t measures of b o t h m e a n NCT a n d / o r a^CT. I n a l l s tudies , p h a n t o m p o s i t i o n was i d e n t i c a l for b o t h the image of interest a n d the b a c k g r o u n d scans w h i c h p r o v i d e d a u t o m a t i c image reg i s t rat i on . W h e n required , image averaging was per formed pr ior t o back -g r o u n d s u b t r a c t i o n . A d d i t i o n a l image process ing a n d analys is steps were requ i red for the i n d i v i d u a l studies c o m p r i s i n g th is thesis (e.g. image f i l t er ing , e x t r a c t i o n of image noise d a t a , measurement of gel v i a l m e a n NCT etc. ) . These techniques, also T u b e voltage ( k V ) T u b e current ( m A ) Slice scan t i m e (s) Slice thickness (mm) F i e l d of v i e w ( c m 2 ) 80, 100, 120, 140 100, 150, 200, 250, 300, 380 0. 8, 1, 2, 3, 4 1, 3, 5, 7, 10 25 x 25, 40 x 40 S t a n d a r d , Soft, L u n g , D e t a i l , Bone , E d g e R e c o n s t r u c t i o n a l g o r i t h m per formed i n M a t L a b , are deta i led i n the respect ive studies, chapters 5, 6 a n d 7. Chapter 5 Gel Composition Studies T h i s chapter provides detai ls of studies per formed to assess a n d bet ter u n d e r s t a n d the effects of gel c o m p o s i t i o n o n the response of p o l y m e r gel to C T read-out a n d o n the u n d e r l y i n g densi ty change. I n i t i a l studies are presented i n sect ion 5.2. These inc lude the design a n d tes t ing of a C T i m a g i n g p h a n t o m o p t i m i z e d for gel v ia ls (section 5.2.1) a n d the es tab l i shment of the r e p r o d u c i b i l i t y of the C T dose response (section 5.2.2). T h e effects of cross - l inker f rac t i on (%C) o n the gel C T dose response are presented i n sect ion 5.3. I n p a r t i c u l a r , response sens i t i v i ty a n d dose range are discussed i n sections 5.3.1 a n d 5.3.2, respectively. T h i s w o r k is fo l lowed i n sect ion 5.4 b y invest igat ions to better u n d e r s t a n d the effect of gel c o m p o s i t i o n o n the i r r a d i a t i o n i n d u c e d densi ty change at the root of the C T dose response. A theore t i ca l m o d e l to describe P A G densi ty change is proposed i n sect ion 5.4.1. A p p l y i n g th is m o d e l to exper imenta l C T a n d R a m a n spectroscopic d a t a , two f u n d a m e n t a l propert ies of the response of P A G dens i ty to dose are discovered. These results are presented a n d discussed i n sections 5.4.2 a n d 5.4.3. M u c h of the w o r k presented i n th is chapter resu l ted i n a research paper p u b l i s h e d i n the i n t e r n a t i o n a l , refereed j o u r n a l Physics in Medicine and Biology [144]. A d d i t i o n a l por t ions of th is w o r k f o rmed p a r t of a second research paper recent ly pub l i shed i n the same j o u r n a l [145]. 5.1 Experimental Details 5.1.1 G e l C o m p o s i t i o n s A l l P A G s used i n these studies were m a n u f a c t u r e d as descr ibed i n sect ion 4.1. T h e weight f rac t i on of t o t a l m o n o m e r (%T) a n d the weight f r a c t i o n of m o n o m e r t h a t is crpssl inker (%C) were var i ed i n order to s t u d y the effects of these parameters o n dose response. T h e p a r t i c u l a r gel f o rmulat i ons invest igated are s u m m a r i z e d i n tab le 5.1. A l l gel batches i n c l u d e d 5% weight f rac t i on of ge la t in . Tab le 5.1: P o l y m e r gel f o rmulat ions used i n the P A G c o m p o s i t i o n a l studies . % T % C 6 0 6 30 6 50 6 70 3 0 3 100 5.1.2 C T I m a g i n g T h e general C T i m a g i n g technique a n d C T scanner ( G E H i S p e e d C T / i ) as descr ibed i n sect ion 4.3 were used for a l l v i a l i m a g i n g . T h e i m a g i n g pro to co l was: 120 k V , 200 m A s , 25 x 25 c m 2 f ie ld of v iew, 1 c m slice th ickness a n d the s t a n d a r d re cons t ruc t i on a l g o r i t h m . T h e 10 v ia l s for each b a t c h of gel were i m a g e d together u s i n g a s tyro foam p h a n t o m . T h i s p h a n t o m , shown i n figure 5.1, was designed to improve image s igna l to noise ra t i o ( S N R ) c o m p a r e d to prev ious gel v i a l C T i m a g i n g techniques [43]. D e t a i l s of the design a n d t es t ing of th is p h a n t o m are discussed below, i n sect ion 5.2.1. F o r each set of gel v ia l s , 16 i d e n t i c a l images were ob ta ined a n d averaged to increase image S N R [39]. I n a d d i t i o n , a set of 10 u n i r r a d i a t e d b a c k g r o u n d gel v ia l s were imaged a n d used to remove artefacts b y b a c k g r o u n d s u b t r a c t i o n [39, 43]. T h e F i g u r e 5.1: T h e p h a n t o m used for C T i m a g i n g gel v ia l s . T h e p h a n t o m design al lows for s imultaneous i m a g i n g of a set of v ia l s a n d uses s tyro foam to improve image S N R key to effective b a c k g r o u n d s u b t r a c t i o n is to have b a c k g r o u n d images w i t h NCT s i m i l a r to the image of interest . It was f ound t h a t ge la t in concentrat i on has the greatest affect on the dens i ty (and therefore NCT) ° f u n i r r a d i a t e d P A G . A s such, for o p t i m u m artefact removal ge la t in concentrat ion s h o u l d be m a t c h e d between the gel of interest a n d the b a c k g r o u n d gel . Since a l l the P A G formulat ions i n these studies conta ined 5 % ge lat in , a single set of b a c k g r o u n d gels was used to per f o rm a l l the b a c k g r o u n d subtrac t ions . These gels were 6 % T , 50 %C P A G s that were inac t ive as they were made i n a n oxygen r i c h env i ronment . F o r each b a t c h of gel , the i r r a d i a t e d set of gels was imaged , removed f r o m the i m a g i n g p h a n t o m a n d replaced w i t h the background v ia ls w i t h o u t m o v i n g the p h a n t o m . T h i s ensured that the l o c a t i o n of the v ia ls was ident i ca l i n the i r r a d i a t e d a n d b a c k g r o u n d gel i m a g i n g sets a n d p r o v i d e d a u t o m a t i c image reg i s t ra t i on . 5.1.3 I m a g e P r o c e s s i n g a n d D a t a A n a l y s i s F o l l o w i n g image averaging a n d b a c k g r o u n d s u b t r a c t i o n , m e a n NCT values were ex-t r a c t e d f r o m the images at regions of interest (21 x 21 pixels ) i n the centre of each v i a l . Va lues of ANCT were ca l cu la ted b y s u b t r a c t i n g NCT measured for the u n i r r a d i -a t e d gel v i a l f r o m a l l other values. Dose responses were p l o t t e d a n d fit w i t h l inear or mono -exponent ia l funct ions u s i n g O r i g i n 7.0 ( O r i g i n L a b C o r p o r a t i o n , N o r t h a m p t o n M A , U S A ) . E r r o r bars i n measured ANCT values were der ived f r om r e p r o d u c i b i l i t y studies (presented i n sect ion 5.2.2, below) to assess the i n t r a a n d i n t e r - b a t c h v a r i -a t i o n i n the C T dose response. T h e error bars o n a l l parameters ca l cu la ted f r o m measured ANCT were der ived f r o m the errors i n the fits us ing s t a n d a r d rules of error p r o p a g a t i o n [189]. 5.1.4 R a m a n S p e c t r o s c o p y R a m a n spectroscopy d a t a is used i n th i s w o r k i n c o n j u n c t i o n w i t h the C T e x p e r i -ments descr ibed above i n order to i l l u m i n a t e features of P A G dens i ty change (section 5.4.2). T h i s is p u b l i s h e d d a t a f r o m exper iments per f o rmed b y a prev ious student i n our lab [86, 191]. I n brief, R a m a n spec t ra were acqu i red o n a B r u k e r R F S F o u r i e r t r a n s f o r m R a m a n spectrometer ( B r u k e r Spec t rosp in , M i l t o n O N , C a n a d a ) u s i n g a 1064 n m N d : Y A G laser o p e r a t i n g at 200 m W . R a m a n spectroscopic features for a c r y l a m i d e (1285 c m - 1 ) , b is (1256 c m - 1 ) a n d p o l y m e r (2936 c m - 1 ) were used to character ize m o n o m e r c o n s u m p t i o n a n d p o l y m e r f o r m a t i o n w i t h dose. F o l l o w i n g b a c k g r o u n d s u b t r a c t i o n , c o r re la t i on was used to character ize the change i n peak intensit ies w i t h dose. T h e reader is referred to references [80, 86, 191] for further detai ls . 5.2 Results and Discussion I: Initial Studies 5.2.1 C T P h a n t o m D e s i g n T h e on ly system used p r e v i o u s l y for C T i m a g i n g v ia l s of p o l y m e r gel p laced the gel v ia ls w i t h i n a water f i l l ed t a n k [43]. T h i s was done i n efforts to reduce image artefacts b y m a t c h i n g gel a n d p h a n t o m densit ies . However , such a sys tem great ly attenuates the x - r a y b e a m a n d , as a result , produces noisy images. T h i s noise increases the u n c e r t a i n t y i n the v i a l r ead ing ( C T J V c t ) a n d c o n t r i b u t e d to poor dose reso lut i on [190]. A s a n a l ternat ive to th is water based p h a n t o m , the poss ib i l i t y of u s i n g s tyro foam for hous ing the gel v ia l s d u r i n g C T i m a g i n g is invest igated i n th is work . T h e effect of C T p h a n t o m m a t e r i a l o n the noise i n gel v i a l measurements was s tud ied b y i m a g i n g three p h a n t o m s of the same d imens i on (29 x 29 c m 2 ) bu t construc ted of different mater ia l s : a c r y l i c , s o l id water a n d s tyro foam. A consistent i m a g i n g pro to co l was used throughout . No ise measurements , O ~ N C T , were m a d e for 10, 21 x 21 p i x e l R O I s , i dent i ca l l y pos i t i oned i n the images of a l l three phantoms . T h i s was done i n order to s imulate the re la t ive a^CT ex t rac ted f r om gel v ia l s i f these mater ia l s were used to house the v ia ls d u r i n g i m a g i n g . 8 F 7 -6 -5 -X V " z B 3 -2 -1 -0 [ _ L _ styrofoam solid water acrylic Phantom material F i g u r e 5.2: Effect of m a t e r i a l hous ing gel v ia l s o n the u n c e r t a i n t y i n measurement of v i a l N C T (a^CT). T h i s uncer ta in ty affects the prec i s ion of dose response curves ex t rac ted f r o m these v i a l readings. F i g u r e 5.2 summarizes the noise levels measured u s i n g the three p h a n t o m mater ia l s . T h e results show that s ty ro f oam prov ides a h i g h l y s ignif icant (~ 11 t imes) improvement i n S N R compared to water . T h e improvement c o m p a r e d to F i g u r e 5.3: R a w C T image of iner t , u n i r r a d i a t e d gel v ia ls o b t a i n e d us ing the s tyro -f oam C T i m a g i n g p h a n t o m s h o w n i n figure 5.1. Not i c e the n o n - u n i f o r m i t y i n NCT w i t h i n the gel v ia ls . a c ry l i c is even greater. B a s e d o n these results , a s tyro foam C T p h a n t o m , as shown i n figure 5.1, was designed to house a l l 10 v ia l s f r om a g iven b a t c h of gel a n d to fit w i t h i n a s m a l l s canning F O V (25 x 25 c m 2 ) . F i g u r e 5.3 shows a n image ob ta ined f r om averaging 64 images of u n i r r a d i a t e d , inert gel v ia ls p laced w i t h i n th is p h a n -t o m . N o n - u n i f o r m i t y i n th is image is apparent w i t h i n the v ia l s . T h e c u p p i n g effect suggests the cause of the n o n - u n i f o r m i t y is b e a m harden ing , as descr ibed i n sect ion 2.4.2. T h e effectiveness of b a c k g r o u n d s u b t r a c t i o n at r e m o v i n g these artefacts was tested by i m a g i n g two sets of gels w i t h different densities a n d s u b t r a c t i n g the images. Prof i les were t h e n ex t rac ted across the v ia l s a n d e x a m i n e d for un i formity . F i g u r e 5.4 compares a n example of profiles o b t a i n e d for b o t h a raw a n d a b a c k g r o u n d sub-t r a c t e d image. Not i c e tha t NCT is u n i f o r m across the v i a l i n the subtrac ted image. T h i s indicates successful artefact remova l a n d val idates the s tyro foam p h a n t o m C T i m a g i n g technique. A n example of a final averaged a n d b a c k g r o u n d subtrac ted i m -age of irradiated gel v ia l s is shown i n figure 5.5. A g a i n , artefacts such as those seen i n figure 5.3 are absent a n d NCT is u n i f o r m w i t h i n each v i a l i n the image. Images such as th is were used to ex trac t the gel dose response d a t a presented i n sections 5.3 a n d 5.4. 20 -r •••• • ' i > i Raw Image Background Subtracted Image 0 -20 --40 i . i . i -20 40 60 80 Pixels F i g u r e 5.4: E x a m p l e of profiles t a k e n across a r a w a n d b a c k g r o u n d subt rac ted C T image of gel v ia ls ob ta ined u s i n g the s tyro foam C T i m a g i n g p h a n t o m . T h e uni for -m i t y i n NCT f ° r the b a c k g r o u n d subt rac ted image val idates the s tyro foam p h a n t o m technique. F i g u r e 5.5: A n example of a f ina l averaged a n d b a c k g r o u n d subtrac ted image of i r r a d i a t e d gel v ia ls o b t a i n e d us ing the p h a n t o m p i c t u r e d i n figure 5.1. N o t i c e tha t NCT is u n i f o r m w i t h i n the v i a l regions 5.2.2 D o s e R e s p o n s e R e p r o d u c i b i l i t y P r e v i o u s w o r k has i l l u s t r a t e d excellent r e p r o d u c i b i l i t y of the C T response of a single P A G dos imeter w h e n repeatedly C T i m a g e d over var ious days [39, 42]. T h i s d e m o n -s t ra ted a h i g h degree of s t a b i l i t y i n the dens i ty change t h a t occurs i n P A G a n d also the a b i l i t y of C T i m a g i n g to prov ide consistent results w h e n i m a g i n g a g iven system. T h i s prev ious w o r k d i d not however address the i n t r a - b a t c h or i n t e r - b a t c h repro -d u c i b i l i t y of P A G C T dose response. T h e charac te r i za t i on of these reproduc ib i l i t i e s is i m p o r t a n t for accurate i n t e r p r e t a t i o n of gel c o m p o s i t i o n studies a n d is per formed here. I n t r a - b a t c h T h e i n t r a - b a t c h r e p r o d u c i b i l i t y of P A G C T dose response assesses the consistency of b o t h the p h a n t o m s used to house the gel v ia l s , s h o w n i n figure 4.2, a n d the v i a l i r r a d i a t i o n technique , descr ibed i n sect ion 4.2.2. A l l v ia l s f r om a single b a t c h of 6 % T , 50 % C gel were i r r a d i a t e d to 8 G y a n d i m a g e d together as descr ibed i n sect ion 5.1.2. T a b l e 5.2 shows the m e a n NCT ex t rac ted f r o m each v i a l . T h e excellent r e p r o d u c i b i l i t y of these results ( s tandard d e v i a t i o n is 0.2 H ) ind icates t h a t b o t h the gel p h a n t o m s a n d i r r a d i a t i o n technique prov ide consistent measurements a n d val idates the use of b o t h . I n t e r - b a t c h Before inves t igat ing different f o rmulat ions of P A G , the r e p r o d u c i b i l i t y of the C T dose response for several batches of the same gel c o m p o s i t i o n needed to be estab-l i shed . T h i s provides error bars on dose response measurements a n d al lows changes i n dose response observed i n later studies to be a t t r i b u t e d to changes i n gel c o m p o -s i t i on . F i g u r e 5.6 shows the dose responses ex t rac ted f r o m four independent batches of P A G (a l l 6 % T , 50 % C ) made over a p e r i o d of several weeks. T h e responses are shown fit to mono -exponent ia l funct ions . F i t t i n g was per fo rmed u s i n g O r i g i n , as T a b l e 5.2: I n t r a - b a t c h r e p r o d u c i b i l i t y of P A G dose response. A l l v ia l s were i r r a d i -a t e d to 8 G y . V i a l NCT (H) 1 10.7 2 10.8 3 10.6 4 10.6 5 10.2 6 10.5 7 10.5 8 10.5 9 10.6 descr ibed i n sect ion 5.1.3. T h e r e p r o d u c i b i l i t y is excel lent. A l l four dose responses were f ound to agree w e l l w i t h i n the uncerta int ies i n the fit parameters . T o reiterate f r o m sect ion 5.1.2, there are several e x p e r i m e n t a l considerat ions requ i red i n order to achieve such excellent r eproduc ib i l i t y . T h e gels must be at s i m i l a r t emperatures w h e n scanned (i.e. c o ld f r o m the refr igerator or at r o o m t e m -perature , as sens i t i v i ty changes by o n l y ~ 0.5 % ° C _ 1 ) a n d the C T scanner s h o u l d be fu l l y w a r m e d u p [39]. I n a d d i t i o n , the gel m a n u f a c t u r i n g technique must be con -sistent. These requirements are not dif f icult to meet a n d i t is i m p o r t a n t to note t h a t th i s is i n m a r k e d contrast to M R I gel d o s i m e t r y for w h i c h h i g h l y reproduc ib le dose responses are more dif f icult to achieve. T h i s is large ly due to the d r a m a t i c effect of i m a g i n g t empera ture o n q u a n t i t a t i v e M R I , ~ 7 % ° C _ 1 [35]. T h e h i g h repro -d u c i b i l i t y of the dose response d e m o n s t r a t e d here is a va luab le result as i t indicates the p o t e n t i a l for C T gel dos imetry to prov ide accurate dose measurements u s i n g a s t a n d a r d c a l i b r a t i o n curve. T h i s w o u l d remove the need to es tab l i sh i n d i v i d u a l c a l i b r a t i o n curves for each b a t c h of dos imeter gel a n d prov ide a p r a c t i c a l advantage over other c o m m o n l y used dos imetry techniques such as f i l m . 12 J i I i I • I • I i I i I i I i I i I i L_ 0 2 4 6 8 10 12 14 16 18 2 0 Dose (Gy) F i g u r e 5.6: C T dose responses measured for four independent batches of i d e n t i c a l p o l y m e r gel ( compos i t i on 6 % T , 50 % C ) . T h e r e p r o d u c i b i l i t y is excel lent: the dose responses a l l agree we l l w i t h i n the errors o n the fits. 5.3 Results and Discussion II: Effect of Cross-linker Frac-tion on Dose Response Increasing % T has been shown to increase the C T dose response of P A G [43]. H o w -ever, the effect of % C has not been p r e v i o u s l y s tud ied . T h i s sect ion provides results w h i c h show the effect of % C o n the gel C T dose response a n d i n p a r t i c u l a r o n the sensi t iv i ty , sect ion 5.3.1, a n d dose range, sect ion 5.3.2. Dos imeter sens i t i v i ty a n d dose range are b o t h i m p o r t a n t c l i n i c a l cons iderat ions a n d , as discussed below, are de termined b y the nature of the dose response. F i g u r e 5.7 shows the C T dose response for P A G s of v a r y i n g % C b u t con-stant 6 % T . N o t e that due to the s o l u b i l i t y l i m i t s of bis , a 100 % C gel c o u l d not be m a n u f a c t u r e d at 6 % T . T h e dose responses c l ear ly v a r y b o t h i n s t rength a n d func-t i o n a l f o rm. T h e 30 a n d 50 % C gels e x h i b i t r e la t ive ly s trong , e x p o n e n t i a l responses a n d the 0 a n d 70 % C gels r espond more w e a k l y a n d l i n e a r l y w i t h dose. T h i s t r e n d is q u a l i t a t i v e l y s i m i l a r t o the effect of % C o n the dose response of P A G observed u s i n g N M R a n d R a m a n spectroscopy [35, 86]. I n these cases the reduced overa l l response of the l inear a n d h i g h l y c ross - l inked systems was a t t r i b u t e d to the di f f i -c u l t y of monomers p e n e t r a t i n g react ive endsites. T h i s is caused b y a h i g h v iscos i ty ( in l inear systems) a n d t i g h t l y f o rmed b e a d i n g ( in h i g h l y c ross - l inked systems) [86]. These same effects are l i k e l y c o n t r i b u t i n g to the responses observed i n figure 5.7, as is discussed further i n sect ion 5.4. T h e effect of % C on the sens i t i v i ty a n d dose range of these responses a n d the p r a c t i c a l i m p l i c a t i o n s of the observed var ia t i ons are presented a n d discussed i n sections 5.3.1 a n d 5.3.2 below. 5.3.1 S e n s i t i v i t y T h e effect of P A G % C o n the sens i t i v i ty of C T dose response is g iven i n figure 5.8. T h e sens i t iv i ty is g iven b y the instantaneous slope of the responses s h o w n i n figure 5.7. T h e l i n e a r l y r espond ing gels have constant sens i t iv i ty over the measured T— 1 —i— 1 —i— 1 —i—•—i—•—i—•—r Dose (Gy) F i g u r e 5.7: T h e effect of % C o n the measured C T dose response of P A G gel dos ime-ters. A l l P A G s are 6 % T . E r r o r bars are der ived f r o m a ± 0.2 H i n t e r - b a t c h repro -d u c i b i l i t y i n NCT measurements . 1.6 1.4 1 1 • 1 • 1 • 1 _ % 0 % C 30 % C /Gy) 1.2 1.0 - • \ 50 % C '_ 70 % C X 0.8 \ . -Sensiti 0.6 -CT 0.4 0.2 ~~ - - ^ 0.0 . 1 . 1 . 1 . 1 . 1 . 1 . 1 . 1 . 1 . 0 2 4 6 8 10 12 14 16 18 20 Dose (Gy) F i g u r e 5.8: T h e effect of P A G % C o n the c a l c u l a t e d sens i t iv i ty of P A G gel dos ime-ters. A l l P A G s are 6 % T dose range (0 to 20 G y ) . T h e sens i t i v i ty of 30 a n d 50 % C P A G s b o t h decrease towards higher doses due to the m o n o - e x p o n e n t i a l na ture of the i r dose responses. A s a result , as shown i n figure 5.8, the most sensit ive % C for C T gel d o s i m e t r y depends o n the dose of interest . T h i s dependence has been i l l u s t r a t e d p r e v i o u s l y u s i n g R a m a n spectroscopy [86] a n d these C T results show a s i m i l a r p a t t e r n : a 30 % C P A G is most sensitive at l ow doses, a 50 % C P A G at a m i d range of doses (~ 4 to 16 G y ) a n d a 0 or 70 % C P A G at h igher doses. T h e i m p l i c a t i o n is t h a t careful select ion of P A G f o r m u l a t i o n is i m p o r t a n t for o p t i m i z i n g gel s ens i t i v i ty for C T read-out w i t h i n a required dose range. A s w i l l be discussed i n sect ion 6.2.7, th i s v a r i a t i o n i n sens i t i v i ty plays a n i m p o r t a n t role i n d e t e r m i n i n g the dose reso lut i on t h a t is achievable us ing different P A G dosimeters . 5.3.2 D o s e R a n g e T h e range of doses over w h i c h a gel c a n operate is another i m p o r t a n t character is t i c for dos imetry . T h e f u n c t i o n a l f o rm a n d sens i t i v i ty of dose response are b o t h factors i n d e t e r m i n i n g this dose range. F o r example , r equ i red use of a l inear dose response or a p a r t i c u l a r gel sens i t iv i ty c a n l i m i t the useable dose range. (Note t h a t l inear dose responses are desirable as they can be used to p e r f o r m re lat ive d o s i m e t r y w i t h o u t c a l i b r a t i n g the dosimeter. ) Consequent ly , f r o m figures 5.7 a n d 5.8, we see t h a t i n C T P A G dos imetry the useful dose range depends o n % C . F o c u s i n g first o n dose response l inear i ty , 0 a n d 70 % C P A G s have l inear dose responses for doses f r o m 0 u p to a least 20 G y . T h e 50 % C P A G has a dose response t h a t can be a p p r o x i m a t e d as l inear u p to ~ 8 G y ( R 2 > 0.988). However , the 30 % C P A G is non- l inear throughout the measured dose range a n d cannot be used to prov ide a l inear dose response. F i g u r e 5.9 shows the C T dose response measured for a n a d d i t i o n a l P A G f o r m u l a t i o n , 3 % T a n d 100 % C , over a n extended dose range. T h i s c o m p o s i t i o n was invest igated as i t has prev ious ly s h o w n cont inued m o n o m e r c o n s u m p t i o n out to ~ 90 G y [86]. A s s h o w n i n figure 5.9, the C T dose response for th is 100 % C P A G is s t i l l increas ing out to 100 G y . A l t h o u g h the response appears s l i gh t ly mono -exponent ia l (x 2 = 0.010), a l inear fit to 100 G y remains p laus ib le (x 2 = 0.023). T h i s extended dose range is i n m a r k e d contrast to the m i d - r a n g e % C P A G s shown i n figure 5.8 for w h i c h dosimeter response is c l ear ly s a t u r a t i n g at m u c h lower doses. A s such, th i s gel f o r m u l a t i o n has p o t e n t i a l for use i n h i g h dose a n d dose range appl i cat ions . A n example is h i g h dose rate brachytherapy . T < 1 • 1 • 1 • r ' i i I i I i I i i _ 0 20 40 60 80 100 Dose (Gy) F i g u r e 5.9: T h e extended C T dose response for a 100 % C , 3 % T P A G . B o t h a mono -exponent ia l (x2 = 0.010) a n d a l inear fit (x 2 = 0.023) to 100 G y , are shown. If a l inear dose response is not requ i red a d d i t i o n a l poss ib i l i t ies exist . F o r example , the 30 % C gel a n d the 50 % C gel at h igher doses (> 8 G y ) m a y be useful i n ce r ta in app l i ca t i ons . However , as s h o w n i n figure 5.8, the sens i t iv i t ies of the C T dose responses of these P A G s decrease w i t h increas ing dose. G i v e n dose reso lut ion depends on sens i t iv i ty (as descr ibed i n sect ion 3.5), th i s w i l l u l t i m a t e l y c o n s t r a i n the useable dose range. F o r example , i f the m i n i m u m requ i red dosimeter sens i t iv i ty is 0.4 H / G y , the 30 a n d 50 % C P A G s w o u l d have m a x i m u m o p e r a t i o n a l dose ranges of ~ 8 a n d 11 G y , respectively. 5.4 Results and Discussion III: P A G Density Change I n th is sect ion the effect of P A G c o m p o s i t i o n o n the r a d i a t i o n i n d u c e d dens i ty change is invest igated i n order to g a i n a bet ter f u n d a m e n t a l u n d e r s t a n d i n g of gel C T dose response. I n sect ion 5.4.1, a theore t i ca l m o d e l is developed to describe the response of P A G dens i ty to dose. B y a p p l y i n g th is m o d e l t o e x p e r i m e n t a l d a t a two fundamenta l propert ies of the r e la t i onsh ip between gel c o m p o s i t i o n a n d the observed densi ty to dose response are discovered. These invest igat ions are presented a n d discussed i n sections 5.4.2 a n d 5.4.3. 5.4.1 T h e o r e t i c a l M o d e l A s discussed i n sect ion 3.4.1, theory suggests t h a t the contrast observed i n C T images of i r r a d i a t e d P A G is due to a dens i ty change t h a t occurs u p o n i r r a d i a t i o n . T h i s has recently been conf i rmed b y independent measurements of P A G dens i ty as a func t i on of dose [140, 148]. T h e fo l l owing presents a n o r i g i n a l m o d e l developed to describe the r a d i a t i o n i n d u c e d dens i ty change. Since, as descr ibed i n 3.2.2, i r r a d i a t i o n of P A G in i t ia tes p o l y m e r i z a t i o n , i t is reasonable to assume that i t is the p o l y m e r i z a t i o n of P A G t h a t causes the dens i ty change. A s such, for a g iven dose the dens i ty change o c c u r r i n g i n P A G (Apgei) can be expressed as a f u n c t i o n of the weight f r a c t i o n of p o l y m e r f o rmed i n the gel a n d a n i n t r i n s i c densi ty change t h a t occurs per weight f rac t i on of m o n o m e r converted to p o l y m e r (Appoiymer). I n other words , the dens i ty change t h a t occurs i n P A G u p o n i r r a d i a t i o n is g iven by Apgel = %PAppoiymer (5.1) where %P is the weight f rac t i on of f o rmed p o l y m e r i n the gel. F o r s impl i c i ty , f r o m th i s po in t o n the t e r m p o l y m e r y i e l d w i l l be used w h e n re ferr ing to %P. T h i s s h o u l d however not be confused w i t h t rue chemica l y i e l d , m o l / J . F o r a P A G sys tem at f u l l p o l y m e r i z a t i o n l i t e ra ture indicates t h a t near ly a l l the avai lable monomer i n the sys tem is c onsumed [80, 196, 197]. T h i s means t h a t %P is 0 for a n u n i r r a d i a t e d gel a n d ~ % T , as g iven i n the i n i t i a l gel c o m p o s i t i o n (%ToGy), for a f u l l y p o l y m e r i z e d gel . It fol lows t h a t for a n y dose %P c a n be expressed as a f u n c t i o n of the f rac t i on of p o l y m e r f o rmed (fp) a n d %Tooy'-%P = %T0Gyfp (5.2) where fp is 0 for a n u n i r r a d i a t e d gel a n d 1 for a f u l l y p o l y m e r i z e d gel. I n a d d i t i o n , fp is r e la ted to the f rac t i on of m o n o m e r i n the gel ( / m ) b y fp = l-fm (5-3) where fm is 1 for a n u n i r r a d i a t e d gel a n d 0 for a f u l l y p o l y m e r i z e d gel . It fol lows f r o m equat ions 5.2 a n d 5.3, t h a t p o l y m e r y i e l d is g iven b y %P = %T0Gy(l - fm). (5.4) C o m b i n i n g equations 5.1 a n d 5.4, the change i n P A G dens i ty t h a t occurs u p o n i r r a d i a t i o n can be expressed as APgel — %ToGy(l — fm)A-Ppolymer- (5-5) T h i s f i n a l equat ion indicates t h a t for a g iven dose the t o t a l dens i ty change that occurs i n P A G , Apgei, depends on the a m o u n t of m o n o m e r i n the i n i t i a l P A G c o m -p o s i t i o n , %ToGy, the f rac t i on of m o n o m e r consumed, (1 — fm), a n d a n i n t r i n s i c dens i ty change t h a t occurs per weight f rac t i on convers ion of monomer to p o l y m e r , Appoiymer- F o r a g iven P A G f o r m u l a t i o n , %Tocy is a constant a n d fm has been shown e x p e r i m e n t a l l y to v a r y w i t h dose i n a m a n n e r t h a t depends o n the p a r t i c u l a r P A G f o r m u l a t i o n [86, 87]. T h e re la t i onsh ip of Appoiymer to dose a n d gel f o r m u l a t i o n is cur rent ly not k n o w n . A s is shown i n the fo l lowing sections, th is m o d e l , together w i t h e x p e r i m e n t a l invest igat ions , i l l u m i n a t e s two i m p o r t a n t f u n d a m e n t a l propert ies of the gel dens i ty response to dose: (1) the dependence of Appoiymer on P A G % C a n d (2) the l inear re la t i onsh ip of Apgei w i t h % T . 5.4.2 E f f e c t o f C r o s s - l i n k e r F r a c t i o n A s ment i oned br ie f ly i n sect ion 5.3, i n p o l y a c r y l a m i d e systems % C affects the t y p e of p o l y m e r f o rmed [86, 198-201]. L o w % C P A G s produce p o l y m e r composed of l ong , l inear , v iscous chains . A t the other extreme, h i g h %C P A G s produce p o l y -mer composed of t i g h t l y w o u n d k n o t s a n d beads. I n the m i d d l e , in termed ia te % C P A G s produce a more web- l ike p o l y m e r s t ruc ture . T h i s v a r i a t i o n i n type of p o l y m e r formed has, for example , mani fested i tsel f i n the d r a m a t i c dependence of fm a n d transverse N M R r e l a x a t i o n rate (R2) o n P A G % C [35, 86]. In tu i t i ve ly , i t w o u l d seem u n l i k e l y that the f o r m a t i o n of such different types of p o l y m e r w o u l d produce the same in t r ins i c dens i ty change o n convers ion of m o n o m e r to p o l y m e r . T h e r e l a -t i onsh ip between Appoiymer a n d % C is invest igated i n th i s sect ion. T h e dens i ty to dose response (Apge[ dose response) for P A G s of v a r y i n g % C (gel batches 2 - 6, tab le 5.1) are s h o w n i n figure 5.10. T h e 0 a n d 100 % C P A G s are 3 % T a n d the r e m a i n i n g gels are 6 % T . These p a r t i c u l a r compos i t i ons were selected to correspond to avai lable R a m a n spectroscopic d a t a . T h e Apgei dose responses were ca l cu la ted f r o m ANCT measurements as descr ibed b y equat i on 3.24 i n chapter 3. F i g u r e 5.10 shows t h a t the sens i t i v i ty a n d shape of the Apgei dose responses depend s trong ly o n P A G f o r m u l a t i o n (also observed i n sect ion 5.3). B a s e d o n the m o d e l th is is expected since, as g iven b y e q u a t i o n 5.1, these curves depend o n b o t h T 1 i • i 1 i 1 i 1 i 1 i 1 i 1 i 1 i 1 r 0 2 4 6 8 10 12 14 16 18 20 Dose (Gy) F i g u r e 5.10: T h e dens i ty change o c c u r r i n g i n P A G as a f u n c t i o n of dose for P A G s of v a r y i n g % C . T h e 0 a n d 100 % C P A G s are 3 % T a n d the r e m a i n i n g P A G s are 6 % T . E r r o r bars are der ived f r o m a ± 0.2 H i n t e r - b a t c h r e p r o d u c i b i l i t y i n NCT measurements . Dose (Gy) F i g u r e 5.11: T h e f rac t i on of m o n o m e r c onsumed w i t h dose as de termined f r om re lat ive R a m a n intensity . T h e 0 a n d 100 % C P A G s are 3 % T a n d the r e m a i n i n g P A G s are 6 % T . T h i s d a t a is ex t rac ted f r om prev ious ly p u b l i s h e d w o r k a n d is shown here w i t h permiss i on f r om the authors [86]. p o l y m e r y i e l d , % P , a n d the in t r ins i c dens i ty change, Appoiymer-T h e %P c o n t r i b u t i o n to A.pgei can be d e t e r m i n e d u s i n g R a m a n spectroscopy since R a m a n d a t a gives a measure of fm. D a t a showing fm, ex t rac ted f r o m R a m a n studies p u b l i s h e d prev ious ly [86], is r eproduced i n f igure 5.11 for c lar i ty . T h e d a t a shown is for i d e n t i c a l gel f ormulat ions as those i n f igure 5.10. I n a d d i t i o n , the d a t a is p l o t t e d over the same dose range for ease of v i s u a l c o m p a r i s o n . U s i n g th is d a t a a n d % T as g iven i n tab le 5.1, %P was ca l cu la ted u s i n g e q u a t i o n 5.4. T h e results are shown i n figure 5.12. T h e i m p o r t a n t feature to not ice i n th i s p lot is t h a t the intermediate % C P A G s (30 a n d 50 % C ) show m u c h greater p o l y m e r y i e l d t h a n do the low (0 % C ) or h i g h % C (70 a n d 100 % C ) P A G s . T h i s t r e n d derives largely f r o m fm, as is seen by c ompar i son of figures 5.11 a n d 5.12, a n d can be exp la ined q u a l i t a t i v e l y by cons ider ing the type of p o l y m e r b e i n g f o rmed . T h e h i g h v iscos i ty 0 2 4 6 8 10 12 14 16 18 20 Dose (Gy) F i g u r e 5.12: P o l y m e r y i e l d (weight f r a c t i o n of p o l y m e r f o rmed i n the gel) , as a f u n c t i o n of dose for P A G s of v a r y i n g % C . T h e 0 a n d 100 % C P A G s are 3 % T a n d the r e m a i n i n g P A G s are 6 % T . of low % C P A G s a n d the t i g h t l y w o u n d knots a n d beads of h i g h % C P A G s b o t h l i m i t the a b i l i t y of monomers to reach ac t ive e n d sites as requ ired to propagate the p o l y m e r i z a t i o n reac t i on [196, 202]. A t in te rmed ia te % C these factors are b o t h reduced a n d p o l y m e r y i e l d increases. T h e effect of % G on &Pp0lymer c a n be i so lated f r om the overa l l P A G den -s i ty change shown i n figure 5.10 b y us ing the p o l y m e r y i e l d d a t a i n figure 5.12 a n d a p p l y i n g the m o d e l (equat ion 5.1). T h e results , shown i n figure 5.13 are v e r y i n -terest ing . T h e y ind i cate that Appoiyrner behaves i n two d i s t inc t ways: the low a n d h i g h % C P A G s exh ib i t a n in t r ins i c dens i ty change a p p r o x i m a t e l y 1.5 t imes greater .E S 3.0 -2.5 2.0 -is * Hi 3 1.0 0.5 I — • — L . • H - I - : . . i - i - i - i -0 % c 30 % C 50 % C 70 % C 100 % C 0 2 4 10 12 14 16 18 20 Dose (Gy) F i g u r e 5.13: T h e effect of P A G % C o n the i n t r i n s i c dens i ty change, Appoiymer, o c c u r r i n g i n i r r a d i a t e d P A G . t h a n do the intermediate % C P A G s . T h e i m p l i c a t i o n is t h a t the i n t r i n s i c dens i ty change that occurs o n p o l y m e r i z a t i o n of l inear or h i g h l y c ross - l inked p o l y m e r is larger t h a n t h a t w h i c h occurs o n f o r m a t i o n of a more web l ike p o l y m e r s t ruc ture . Since the increase i n p o l y m e r gel dens i ty w i t h dose has been a t t r i b u t e d to a vo lume decrease (as opposed to mass increase) [43], th i s d a t a shows that gel shr inkage per u n i t p o l y m e r y i e l d is greatest for l inear a n d h i g h l y cross l inked systems. A n interest ing observat ion is t h a t these Appoiymer results exh ib i t the oppo -site t r e n d to the p o l y m e r y i e l d results t h a t are shown i n figure 5.12. It appears t h a t Appoiymer a n d p o l y m e r y i e l d are inverse ly re lated : a h i g h y i e l d p o l y m e r sys-t e m is l ike ly to exh ib i t a lower i n t r i n s i c dens i ty change t h a n a lower y i e l d p o l y m e r sys tem. A s was discovered above, the l ong l inear cha ined p o l y m e r of 0 % C P A G s a n d the t i g h t l y beaded a n d k n o t t e d p o l y m e r of h i g h % C P A G s are more dense t h a n the web- l ike p o l y m e r formed at in te rmed ia te values of % C . It is possible that th is increased dens i ty contr ibutes to the low y i e l d observed for the low a n d h i g h l y cross-l i n k e d P A G s , e x p l a i n i n g i n p a r t the inverse re la t i onsh ip . I n terms of o p t i m i z a t i o n of p o l y m e r gel for h i g h sens i t i v i ty to C T i m a g i n g , th is result means that %P a n d App0iymer ( equat ion 5.1) cannot be m a x i m i z e d s imultaneously . However , as shown by figure 5.10 in termed ia te % C P A G s show the greatest Apgei dose responses. T h i s means that %P is more d o m i n a n t t h a n Appoiymer i n d e t e r m i n i n g overa l l P A G den -s i ty change. A s such, th is w o r k suggests t h a t future efforts to increase the sens i t i v i ty of C T gel dos imetry s h o u l d focus o n systems w i t h h i g h p o l y m e r y i e l d ra ther t h a n h i g h in t r ins i c dens i ty change. 5.4.3 E f f e c t o f T o t a l M o n o m e r F r a c t i o n Apgei dose responses were measured for 0 % C P A G s at b o t h 3 a n d 6 % T (batches 1 a n d 5, table 5.1). T h e results are c o m p a r e d i n figure 5.14. T h e responses are l inear for b o t h P A G s : 0.112 ± 0.002 k g r n ^ G y " 1 ( R 2 = 0.9903) a n d 0.226 ± 0.006 k g m _ 3 G y _ 1 ( R 2 = 0.9862) for the 3 a n d 6 % T P A G s , respectively. T h i s ind icates , at least for these 0 % C gels, tha t the f u n c t i o n a l f o r m of the Apgei dose response is independent of % T . I n a d d i t i o n , the Apgei dose response increases as % T increases. T h i s result q u a l i t a t i v e l y agrees w i t h s canning e lectron m i c r o g r a p h observations of increased concentra t i on of s t r u c t u r a l mass per u n i t area w i t h greater m o n o m e r con-centrat ions [201]. A more q u a n t i t a t i v e analys is i l l u m i n a t e s the fact t h a t the response of the 6 % T P A G is , w i t h i n uncerta inty , twice t h a t of the 3 % T P A G . T h i s means that for these gels Apgei is l inear w i t h % T . U s i n g the m o d e l (equat ion 5.5) th is leads to the interest ing i m p l i c a t i o n t h a t b o t h fm a n d Appoiymer are independent of % T . A n o t h e r group s t u d y i n g the x - r a y C T read-out of p o l y m e r gel dosimeters has invest igated the effect of % T o n the C T dose response of 50 % C P A G s [43]. T h e y f ound that the sens i t iv i ty of the C T dose response, as measured i n the quas i - l inear low dose region, increased w i t h increas ing % T . O n close i n s p e c t i o n of the i r d a t a , i t is clear tha t the 50 % C P A G exhib i t s the same l inear re la t i onsh ip w i t h % T as was E t 2 1 --i— 1—i—•—r - i — i i — i — i — • 3 % T • 6 % T 0.226+/-0.006 kg/m3/Gy \ - " V 0.112+/-0.002kg/m7Gy Dose (Gy) 20 F i g u r e 5.14: T h e effect of % T o n the dens i ty dose response of 0 % C P A G s . T h e sens i t iv i ty us ing 6 % T is twice t h a t for 3 % T , w i t h i n u n c e r t a i n t y i n the f its . observed for the 0 % C P A G . T o i l l u s t r a t e th i s c learly, the i r p u b l i s h e d sens i t i v i ty d a t a (converted to densi ty us ing equat i on 3.24) is p l o t t e d against % T i n figure 5.15. T h i s is done w i t h permiss i on f r o m the correspond ing a u t h o r , C . B a l d o c k . T h e result is h i g h l y l inear ( R 2 = 0.9933). A p g e ; is therefore s h o w n e x p e r i m e n t a l l y to be l inear w i t h % T for b o t h 0 a n d 50 % C P A G s , two systems t h a t f o r m d i s t i n c t l y different types of p o l y m e r . T h i s confirms the m o d e l proposed here a n d suggests t h a t i n P A G systems fm a n d Appoiymer are governed b y % C alone, at least for the range of m o n o m e r concentrat ions e x a m i n e d here (2 to 12 % T ) . I n p r a c t i c a l t e rms , th i s result indicates t h a t for m a x i m u m sens i t iv i ty of p o l y m e r gel to C T s c a n n i n g % T s h o u l d be m a x i m i z e d w i t h i n the constra int of p r o d u c i n g a stable , r eproduc ib l e sys tem. 2 4 6 8 10 12 PAG % T F i g u r e 5.15: T h e dens i ty change per u n i t dose as a f u n c t i o n of % T for a 50 % C P A G . These results are ex t rac ted f r o m d a t a p u b l i s h e d p r e v i o u s l y a n d are shown w i t h permiss i on f r o m C . B a l d o c k [43]. 5.5 Chapter Summary These c o m p o s i t i o n a l studies have p r o v i d e d va luab le knowledge about the C T dose response of P A G dosimeters . E x c e l l e n t dose response r e p r o d u c i b i l i t y has been es-t a b l i s h e d a n d v a r i a t i o n i n the sens i t iv i ty a n d dose range of the response for different gel composi t ions , i n p a r t i c u l a r for gels w i t h v a r y i n g % C , have been demonstrated . I n a d d i t i o n , i n p e r f o r m i n g th i s e x p e r i m e n t a l w o r k a m e t h o d of i m a g i n g gel v ia ls was developed that d r a m a t i c a l l y i m p r o v e d s igna l t o noise ra t i o c o m p a r e d to p r e v i -ous methods . F i n a l l y , a m o d e l was developed to describe gel dens i ty change as a f u n c t i o n of p o l y m e r y i e l d a n d a n i n t r i n s i c dens i ty change t h a t occurs o n convers ion of monomer to po lymer . A p p l y i n g th is m o d e l to e x p e r i m e n t a l d a t a , f u n d a m e n t a l propert ies of the gel dens i ty to dose response were discovered. I n s u m m a r y , these c o m p o s i t i o n a l studies p r o d u c e d a n i m p r o v e d m e t h o d of i m a g i n g gel v ia l s , have s ig -n i f i cant ly fur thered the b o d y of f u n d a m e n t a l knowledge about p o l y m e r gel dose response a n d have i l l u m i n a t e d p o t e n t i a l avenues for future w o r k i n deve loping gel f o rmulat ions w i t h i m p r o v e d sens i t iv i ty to C T i m a g i n g . A s discussed i n sect ion 3.5, gel sens i t iv i ty is one i m p o r t a n t factor i n d e t e r m i n i n g dose reso lut ion . A n o t h e r fac-tor , image noise, is invest igated i n the fo l l owing chapter , chapter 6. Chapter 6 Noise and Dose Resolution Studies T h i s chapter describes a series of studies designed to de termine strategies for m i n i -m i z i n g image noise a n d m a x i m i z i n g dose reso lu t i on i n C T gel dos imetry . T h e results prov ide d a t a o n system performance t h a t c o u l d serve as a useful reference for i m p l e -m e n t a t i o n of C T gel dos imetry i n the c l i n i c a l env iroment . T h e effects of p h a n t o m design, C T i m a g i n g technique a n d image voxe l size o n image noise are quant i f ied . These results are c ombined w i t h the P A G sens i t i v i ty results f r om chapter 5 to c a l c u -late achievable dose reso lut ion for C T P A G dos imetry . A s t u d y of the re la t i onsh ip between i m a g i n g t i m e a n d scan technique is used to c o n s t r a i n the dose reso lut ion to tha t achievable w i t h i n a 1 hour i m a g i n g t i m e . Image u n i f o r m i t y a n d the a b i l i t y to predic t image noise for a g iven s cann ing p r o t o c o l are also addressed. T h e re la t i onsh ip between image noise a n d the sys tem parameters s tud ied here c a n i n general be der ived f r om theory. T h i s is due to the pred i c tab le w a y m a n y parameters affect the t o t a l number of photons counted (N) a n d hence noise, as g iven b y equat i on 2.10. F o r example , N increases l i n e a r l y w i t h t u b e current , slice scan t i m e a n d slice thickness . Therefore , increas ing these parameters s h o u l d prov ide a 1/y/N r e d u c t i o n i n image noise. I n a d d i t i o n , based o n the k n o w n effect of a t tenuator thickness o n b e a m in tens i ty g iven b y equat i on 1.6, r e d u c i n g p h a n t o m d iameter is expected to prov ide a n e x p o n e n t i a l r e d u c t i o n i n image noise. T h e results presented here are consistent w i t h theore t i ca l pred ic t ions a n d va l idate p r a c t i c a l strategies for m i n i m i z i n g noise i n C T gel dos imetry . A s we l l , the measured d a t a serves as useful reference i n f o r m a t i o n i n d i c a t i n g the range of noise levels tha t can be expected for gel dos imetry a n d a l l o w i n g p r e d i c t i o n of the achievable dose reso lut i on for a range of voxel sizes a n d i m a g i n g t i m e constra ints . E x p e r i m e n t a l detai ls specific to these studies are g iven i n sect ion 6.1. T h e effects of p h a n t o m design, scan technique a n d voxe l size o n image noise are s tud ied i n sections 6.2.1 t h r o u g h 6.2.3. T h i s w o r k leads to the es tab l i shment of a m e t h o d of p r e d i c t i n g image noise for a n y g iven C T i m a g i n g pro toco l , presented i n sect ion 6.2.4. Image u n i f o r m i t y is assessed, i n the context of noise levels i n gel dos imetry , i n sect ion 6.2.5. T h e effect of s cann ing pro to co l o n i m a g i n g t i m e is establ ished i n sect ion 6.2.6. B a s e d o n these results a n d those of chapter 5, sect ion 6.2.7 presents the achievable dose reso lut ion for C T P A G gel d o s i m e t r y systems g iven voxel size a n d i m a g i n g t i m e constra ints . F i n a l l y , i n sect ion 6.2.8 a c o m p a r i s o n is made between C T gel dos imetry per formance as de termined i n th is w o r k a n d the per formance of M R I a n d O C T gel dos imetry techniques as presented i n recent l i t e ra ture . T h e w o r k presented i n th i s chapter forms a large p o r t i o n of a paper c u r r e n t l y under review by the i n t e r n a t i o n a l j o u r n a l Physics in Medicine and Biology. 6.1 Experimental Details T h e exper iments i n these studies invo lved C T i m a g i n g water f i l l ed phantoms . Spe-cific i m a g i n g methods var i ed for each inves t igat i on a n d detai ls are p r o v i d e d here a long w i t h image processing a n d d a t a analys is procedures . T h e p o l y m e r gel dose response i n f o r m a t i o n used i n the f ina l c a l c u l a t i o n of achievable dose reso lut i on (sec-t i o n 6.2.8) is f r o m the results of the P A G c o m p o s i t i o n study, presented i n chapter 5. E x p e r i m e n t a l detai ls re lated to gel p r o d u c t i o n , i r r a d i a t i o n a n d i m a g i n g are p r o v i d e d i n tha t chapter . 6.1.1 C T I m a g i n g C T i m a g i n g for a l l exper iments was per fo rmed us ing a G E H i S p e e d C T / i C T scanner ( G E M e d i c a l Systems) , descr ibed i n d e t a i l i n chapter 4. T h i s C T scanner is a single slice, 3rd generat ion , ro tate - ro tate mach ine a n d is s i m i l a r to the types of C T scanners c u r r e n t l y avai lable i n most cancer hosp i ta ls . Some measurements conta ined i n th i s s t u d y (such as the effects of C T i m a g i n g p r o t o c o l on image noise a n d i m a g i n g t ime) w i l l depend o n the p a r t i c u l a r make a n d m o d e l of C T scanner. T h i s is exempl i f i ed b y a C T scanner dose survey per formed b y I m P A C T ( L o n d o n U K ) w h i c h shows a s t a n d a r d d e v i a t i o n of ~ 2 0 % i n C T dose across models of single slice C T scanners u s i n g a constant s canning p r o t o c o l [203]. However , a l t h o u g h measured abso lute noise values w i l l v a r y w i t h C T system, the overa l l t rends shown i n th is w o r k w i l l be representative of other single slice machines a n d the o p t i m i z a t i o n strategies proposed w i l l r e m a i n relevant. T h e i n t r o d u c t i o n of m u l t i - s l i c e C T scanners in to r a d i a t i o n t h e r a p y pract i ce w i l l prov ide a new level of per formance ( in p a r t i c u l a r a s igni f icant improvement i n the noise level achievable i n a short i m a g i n g t i m e is predicted) a n d is a top i c for future work , as descr ibed i n chapter 8. T h e detai ls of C T i m a g i n g w h i c h are specific to each of the invest igat ions i n th is s tudy are p r o v i d e d below. I m a g i n g p r o t o c o l exper iments T h e effects of C T i m a g i n g p r o t o c o l o n image noise were s tud ied u s i n g a c y l i n d r i c a l water f i l led p h a n t o m . T h e p h a n t o m was ~ 12.5 c m i n d iameter to m i m i c a t y p i c a l size for a gel dos imetry p h a n t o m . T h e dependence of image noise o n each avai lable C T i m a g i n g parameter was measured i n d i v i d u a l l y . T h i s was achieved b y o b t a i n i n g images of the water p h a n t o m u s i n g scan protoco ls t h a t independent ly v a r i e d each scan parameter l i s ted i n tab le 6.1 f r o m a reference pro to co l (bo lded i n tab le 6.1). A l l i m a g i n g used the smallest avai lable field of v i ew (25 x 25 c m 2 ) a n d p i x e l size T a b l e 6.1: S c a n parameters used to test the effects of scan p r o t o c o l o n image noise. P a r a m e t e r s i n b o l d face comprise the reference protoco l . R e c o n s t r u c t i o n a l g o r i t h m names are s tandards for G E C T scanners. S c a n parameter (units) Va lues used T u b e voltage ( k V ) T u b e current ( m A ) Sl ice scan t i m e (s) Slice thickness (mm) F i e l d of v i e w ( cm 2 ) 80, 100, 120, 1 4 0 100, 150, 2 0 0 , 250, 300, 380 0. 8, 1, 2, 3, 4 1, 3, 5, 7, 10 2 5 x 2 5 0 . 5 , 1, 1.5, 2, 2.5, 3, 5 P i x e l d i m e n s i o n (mm) R e c o n s t r u c t i o n a l g o r i t h m S t a n d a r d , Soft , L u n g , D e t a i l , B o n e , E d g e was var i ed p o s t - i m a g i n g us ing M a t L a b as descr ibed i n sect ion 6.1.2. N o t e t h a t for each set of scan parameters two images were o b t a i n e d i n order to remove artefacts b y b a c k g r o u n d s u b t r a c t i o n pr ior to m a k i n g noise measurements . P h a n t o m design experiments T h e effect of p h a n t o m m a t e r i a l o n image noise was invest igated i n the context of v i a l i m a g i n g for dose response c a l i b r a t i o n i n chapter 5. T h e effect of p h a n t o m size is invest igated here us ing five glass (2 m m t h i c k P y r e x ) water f i l l ed p h a n t o m s w i t h d iameters of 6, 8, 10, 13 a n d 15.5 c m . P h a n t o m sizes were selected to cover a range relevant for 3 D ver i f i cat ion of c l i n i c a l r a d i a t i o n therapy t reatments . T h e reference pro toco l , b o l d e d i n tab le 6.1, was used for C T i m a g i n g a l l phantoms . I m a g i n g t i m e experiments I m a g i n g t i m e was measured by record ing the average t i m e requ i red to acquire a f i n a l image of a single sl ice, i n c l u d i n g image averaging requirements . F o r m u l t i - s l i c e i m a g i n g , images of several slices were o b t a i n e d sequent ia l ly a n d the average t i m e required to o b t a i n a single slice de te rmined . T h i s was done i n order to inc lude the effect of x - r a y tube heat ing o n i m a g i n g t i m e . T h e effect of i m a g i n g p r o t o c o l on i m a g i n g t i m e was de termined by v a r y i n g the t u b e current a n d n u m b e r of averages i n order to achieve different loads on the x - r a y tube . D e t e r m i n i n g i m a g i n g t i m e per slice i n th is manner al lows for easy c a l c u l a t i o n o f the t o t a l t i m e requ i red for s c a n n i n g the vo lume of a g iven p o l y m e r gel. 6.1.2 I m a g e P r o c e s s i n g a n d D a t a A n a l y s i s Image averaging, b a c k g r o u n d s u b t r a c t i o n a n d image analyses were per fo rmed us ing M a t L a b . B a c k g r o u n d s u b t r a c t i o n was p a r t i c u l a r l y c r i t i c a l to ensure remova l of artefacts tha t might c ontaminate t r u e noise measurements . F o r s t u d y i n g the effect of C T i m a g i n g pro toco l a n d p h a n t o m size o n image noise, the s t a n d a r d d e v i a t i o n i n C T n u m b e r {O~NCT) w a s ex t rac ted f r o m i d e n t i c a l 50 x 50 p i x e l regions of interest (ROIs ) i n the centre of the images. U n i f o r m i t y of C T i m a g i n g was tested b y e x t r a c t i n g m e a n C T number (NCT) a n d CJNCT f r o m 36, 25 x 25 p i x e l R O I s i n the same p h a n t o m used for the C T i m a g i n g pro to co l exper iments . A s s tated above (section 6.1.1) a l l C T i m a g i n g was per fo rmed at the smallest avai lable f ield of v i ew , 25 x 25 c m 2 . G i v e n a 512 x 512 m a t r i x size, th i s provides a p i x e l d imens ion of ~ 0.5 m m . T h e effect of p i x e l size o n image noise was e x a m -i n e d b y increas ing p i x e l d i m e n s i o n u p to 5 m m , as g iven i n tab le 6.1, p o s t - i m a g i n g u s i n g M a t L a b . Noise measurements were o b t a i n e d f r o m each image , as above, by e x t r a c t i n g <7NCT from ident i ca l 50 x 50 p i x e l R O I s i n the centre of the images. 6.2 Results and Discussion 6.2.1 P h a n t o m D e s i g n Image noise a n d artefacts present i n C T images of p o l y m e r gel are b o t h affected b y the p h a n t o m used and , as such, p h a n t o m design is a n i m p o r t a n t factor i n o p t i m i z i n g C T i m a g i n g as a read-out m e t h o d for gel dos imetry . P h a n t o m m a t e r i a l was shown to have a d r a m a t i c effect o n image noise i n chapter 5. P h a n t o m size also affects image noise a n d is invest igated here. T h e e x p e r i m e n t a l results measured for five s i m i l a r p h a n t o m s of v a r y i n g d i a m -eter are shown i n figure 6.1. T h e measured increase i n G N C T w i t h p h a n t o m size fits we l l to the theoret i ca l exponent ia l dependence. T h i s effect is the result of increased p h o t o n a t t e n u a t i o n b y larger p h a n t o m s a n d the ensuing r educ t i on i n the n u m b e r of photons counted a n d hence increase i n q u a n t u m noise [47]. Since image noise d i r e c t l y i m p a c t s dose reso lut ion , C T gel d o s i m e t r y p h a n t o m s s h o u l d be designed so that the i m a g e d p h a n t o m is as s m a l l as p r a c t i c a l for a g iven a p p l i c a t i o n . D u e to the exponent ia l na ture of the effect even s m a l l reduct ions i n p h a n t o m size m a y prov ide a measurable improvement i n image noise. I n l ight of th i s fact, a n o p t i m u m gel sys tem c o u l d be designed so that the gel is i r r a d i a t e d i n a larger p h a n t o m b u t is C T i m a g e d alone, removed f r om th is larger p h a n t o m , i n order to reduce image noise. 10 12 14 Phantom Diameter (cm) F i g u r e 6.1: T h e effect of p h a n t o m size o n C T image noise. T h e fit is exponent ia l . 6.2.2 C T I m a g i n g T e c h n i q u e C T i m a g i n g technique is a n i m p o r t a n t cons iderat i on for m i n i m i z i n g noise i n C T gel dosimetry . I n p a r t i c u l a r , i m a g i n g technique parameters affect the n u m b e r of photons counted (N) a n d hence, as discussed i n sect ion 2.4.1, the amount of q u a n t u m noise degrading the images [47]. These effects are invest igated here, a long w i t h the effect of choice of r e cons t ruc t i on a l g o r i t h m on image noise. These results are i m p o r t a n t as they serve to prov ide guidel ines for select ing a n o p t i m u m gel C T i m a g i n g p r o t o c o l w h e n c o m p e t i n g factors such as i m a g i n g t i m e (discussed i n sect ion 6.2.6) m u s t be considered. Increasing x - r a y t u b e voltage (kV), x - r a y t u b e current (mA) a n d slice scan t i m e (s) a l l result i n a decrease i n image noise (O~NCT) a s l s s n o w n m f igure 6.2. Bes t fits to the d a t a ind i ca te t h a t t u b e current a n d slice scan t i m e decrease image noise b y (mA)"0-57 a n d ( s ) - 0 4 7 , respectively. Since increas ing t u b e current a n d slice scan t i m e each l i n e a r l y increase the number of photons inc ident o n the p h a n t o m , these results are quite consistent w i t h the theoret i ca l 1/V~N r e d u c t i o n i n image noise based on p h o t o n c o u n t i n g s tat is t i cs [47]. N o t e t h a t these re lat ionships to image noise are ident i ca l to t h a t of image averaging , cons is tent ly used i n C T gel d o s i m e t r y as a n a d d i t i o n a l noise r e d u c t i o n t oo l [39, 43]. Image averaging reduces noise b y 1/y/NAX, where NAX is the number of averages. T h e re la t i onsh ip between tube voltage a n d image noise is more complex since t u b e vol tage affects b e a m energy i n a d d i t i o n to TV. T h e results shown i n figure 6.2 i n d i c a t e that increas ing t u b e vol tage decreases image noise b y ~ (kV)~1A. T h i s means t h a t a change i n t u b e voltage has a greater affect o n image noise t h a n does a n equa l percent change i n t u b e current or slice scan t i m e . F o r example , d o u b l i n g t u b e vol tage w o u l d produce a n expected ~ 6 0 % reduc t i on i n image noise whereas d o u b l i n g t u b e current o n l y ~ 3 0 % . Since the l o a d o n the x - r a y t u b e d u r i n g s c a n n i n g is affected equa l ly b y a l l three of these parameters (kJ = kV x mA x s ) , these results ind i cate t h a t increas ing tube voltage is the most efficient means of r e d u c i n g image noise i n a C T s cann ing 1 ' 1 1 r 70 80 90 100 110 120 130 140 150 Tube voltage (kV) i . i . i • i • i • i • i • i • i • i 0 50 100 150 200 250 300 350 400 450 Tube current (mA) i i i — i — i — . — i — i — i — i — i — i — i — i — i — i — i 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Slice scan time (s) F i g u r e 6.2: Effects of scan technique (x -ray t u b e voltage, current a n d slice scan t ime) o n C T image noise. N o t e t h a t different x -ax i s are p r o v i d e d for the three parameters . * Tube voltage (kV) • Tube current (mA) Slice scan time (s) X b 3 -2 20 40 60 80 100 120 X-ray tube load (kJ) F i g u r e 6.3: Image noise measured at a range of x - r a y t u b e loads achieved b y changing each of t u b e voltage, current a n d slice scan t i m e i n d i v i d u a l l y f r o m the s t a n d a r d set of reference parameters (given i n table 6.1). E a c h parameter was var i ed over its f u l l opera t i on range as g iven i n tab le 6.1. T h e curve ob ta ined b y chang ing t u b e voltage is c lear ly m u c h steeper t h a n that o b t a i n e d by chang ing current or slice scan t ime . T h i s indicates tha t increas ing tube voltage is the most efficient means of lowering image noise t h r o u g h choice of scan pro toco l . p ro toco l . T h i s is emphas ized i n figure 6.3 w h i c h shows image noise measured for a range of t u b e loads ob ta ined b y independent ly v a r y i n g each of tube voltage, current a n d slice scan t i m e f r om the reference values g iven i n tab le 6.1. T h i s impl ies tha t m a x i m i z i n g t u b e voltage s h o u l d be the p r i o r i t y w h e n o p t i m i z i n g a C T i m a g i n g pro to co l for low noise but h i g h efficiency C T gel dos imetry . Several reconstruc t i on a lgor i thms are avai lable on C T scanners i n order to a l l ow users to h igh l ight features of interest i n the images. A s descr ibed i n sect ion 2.3.3, these a lgor i thms frequently employ filters to , for example , enhance contrast or de ta i l i n the images. T h e result , as shown i n figure 6.4, is a n enormous v a r i a t i o n i n image noise o b t a i n e d by s i m p l y chang ing recons t ruc t i on a l g o r i t h m . T h e algo-r i t h m s tested here are those avai lable o n the G E H i S p e e d C T / i scanner (see tab le 6.1). A l g o r i t h m detai ls are p r o p r i e t a r y a n d specific a l go r i thms w i l l differ w i t h s can-ner manufacturer . However a s i m i l a r re lat ive v a r i a t i o n c a n be expected for other scanners a n d the general conclusions d r a w n here w i l l r e m a i n relevant. These results demonstrate t h a t reconstruc t i on a l g o r i t h m has the greatest i m p a c t o n image noise of a n y scan technique parameter a n d , as such , choice of a n a p p r o p r i a t e re cons t ruc t i on a l g o r i t h m is c r i t i c a l for o p t i m i z e d C T gel dos imetry . T h e low contrast reso lut ion a l g o r i t h m (e.g. S O F T b y G E ) provides the lowest noise. However , since these a lgo-r i t h m s suppress h i g h frequency i n f o r m a t i o n to reduce noise, i f r e cord ing very h i g h s p a t i a l de ta i l is more c r i t i c a l t h a n h i g h dose reso lut ion , a s t a n d a r d a l g o r i t h m (e.g. S T A N D A R D by G E ) is recommended . A l g o r i t h m s spec i f i ca l ly designed to h igh l ight edges a n d fine s t ruc ture i n l u n g a n d bone s h o u l d be avoided. A deta i l ed e x a m i n a -t i o n of different strategies for pos t - re cons t ruc t i on filtering of C T images to further reduce image noise is presented i n chapter 7 6.2.3 V o x e l D i m e n s i o n s C T i m a g i n g can prov ide images w i t h v e r y s m a l l voxe l d imensions . S c a n n i n g w i t h a s m a l l (25 x 25 c m 2 ) field of v iew, the 512 x 512 m a t r i x size of C T images provides images w i t h a p i x e l d i m e n s i o n of less t h a n 0.5 m m . I n the out of image p lane d imens i on (slice separat ion) , t r a d i t i o n a l single slice C T scanners can t y p i c a l l y prov ide 1 m m reso lut ion a n d some new m u l t i - s l i c e machines can do even better . S m a l l voxe l size i n C T i m a g i n g does however present a n inherent compromise w i t h increased image noise. T h i s is due to decreased p h o t o n c o u n t i n g stat ist ics w i t h decreased voxel size a n d appl ies to b o t h p i x e l size w i t h i n the image p lane (x,y) a n d , w h e n slice thickness equals sl ice separat i on (as is t y p i c a l i n C T i m a g i n g ) , the d istance between image planes (z). T h e measured effects of b o t h p i x e l d i m e n s i o n a n d slice thickness o n image noise are shown i n figure 6.5. T h e re la t i onsh ip between image noise a n d slice t h i c k -I 1 •••I jppwp mmmi • • • • 111111111 • • • I j Li 1 L_i 1 L J 1 L 1 . J 1 L SOFT STANDARD DETAIL BONE LUNG EDGE Reconstruction Algorithm F i g u r e 6.4: T h e effect of choice of r e cons t ruc t i on a l g o r i t h m o n noise i n the r e s u l t i n g C T image. T h e reconstruc t i on a l g o r i t h m s tested are s tandards for G E C T scanners. ness (h) is f ound to be (7NCT oc / i - 0 5 . A l l else equa l , the number of photons counted (N) is expected to increase l i n e a r l y w i t h slice thickness . A s such, th is measured result agrees w i t h the l / V ^ / V r e d u c t i o n i n image noise pred i c t ed b y theory (section 2.4.1). T h e effect of p i x e l d imens i on o n noise is measured to be greater t h a n t h a t of slice thickness a n d , as is shown i n figure 6.5, is descr ibed we l l by a m o n o - e x p o n e n t i a l (x 2 = 0.013). T h e r e is disagreement i n the l i t e ra ture as to the p r e d i c t e d behav iour of image noise w i t h p i x e l d i m e n s i o n (w) i n C T i m a g i n g . M o s t authors suggest a 1/Vw3 r e la t i onsh ip w i t h image noise [57, 204, 205]. However , others suggest t h a t a smal ler effect is to be expected , such as O~NCT OC l/y/w [53]. W h e n fit to th is func-t i o n a l f o rm (O~NCT oc ly/vF), th i s s t u d y ind icates a re la t i onsh ip m i d w a y between these predic t ions : CFNCT OC y/w1-3. However , the fit to th is f u n c t i o n (%2 = 0.098) is not near ly as good as the exponent ia l fit t h a t is shown i n figure 6.5. These results have valuable i m p l i c a t i o n s for o p t i m u m c l i n i c a l i m p l e m e n t a t i o n of C T gel dos imetry since i n m a n y s i tuat i ons a 0.5 m m voxe l d i m e n s i o n m a y prove 25 20 £ 15 rj z b 10 10 Slice thickness (mm) 2 3 Pixel dimension (mm) F i g u r e 6.5: T h e effects of C T image voxel d imensions o n image noise. T h e in -p lane d imens i on (p ixe l size) has a greater affect o n image noise t h a n does the d i m e n s i o n between i m a g i n g planes (slice thickness a n d separat ion) . unnecessary. T h i s w o u l d occur, for example , w h e n us ing gel dos imetry to va l idate t reatment p l a n n i n g ca lcu lat ions per formed o n a c a l c u l a t i o n g r i d that has a 1.25 or 2.5 m m reso lut ion . A s these results have shown, image noise can be i m p r o v e d by ~ 60 % b y increas ing the voxel d imens i on i n the measured gel dos imetry results to 2 m m . T h i s demonstrates , quant i ta t ive ly , the compromise between voxe l size a n d image noise i n C T gel dos imetry a n d h igh l ights the i m p o r t a n c e of o p t i m i z i n g voxe l size based o n the requirements of a g iven a p p l i c a t i o n . 6.2.4 P r e d i c t i n g I m a g e N o i s e F o r ease of reference, the quant i ta t i ve effects o n image noise of a l l C T i m a g i n g fac-tors invest igated i n sections 6.2.1, 6.2.2 a n d 6.2.3 are s u m m a r i z e d i n tab le 6.2. A n e x c i t i n g result for i m p l e m e n t a t i o n of C T gel d o s i m e t r y is tha t a l l the parameters l i s t ed i n tab le 6.1 affect image noise independent ly of one another . T h i s was d e m o n -s t ra ted b y v a r y i n g a l l parameters i n pa i rs , m e a s u r i n g the resu l t ing image noise a n d c o m p a r i n g these results to ca l cu la ted pred i c t i ons based o n the re lat ionships s u m -m a r i z e d i n tab le 6.2. Tab le 6.3 provides several examples u s i n g low, m e d i u m a n d h i g h noise i m a g i n g techniques. T h e results show t h a t measured a n d pred i c t ed image noise agree very w e l l so t h a t , after m e a s u r i n g image noise for a reference i m a g i n g pro toco l , image noise can be pred i c ted for a n y other i m a g i n g pro toco l . T h e p r a c t i c a l i m p l i c a t i o n is t h a t , g iven a k n o w n gel C T dose response, th is feature al lows for se-l ec t i on of a p a r t i c u l a r C T i m a g i n g p r o t o c o l i n order to meet specific dose reso lut i on requirements . T a b l e 6.2: A s u m m a r y of the measured q u a n t i t a t i v e effects of p h a n t o m size, C T s c a n n i n g technique factors a n d voxel size o n C T image noise. F a c t o r affecting image noise (symbol ) R e l a t i o n s h i p to image noise (CFNCT) P h a n t o m diameter (d) aNCT °^ E < I T u b e voltage ( k V ) (TNCT oc (kV)~1A T u b e current ( m A ) ONCT a ( " i ^ 4 ) - 0 ' 5 Slice scan t i m e (s) a N c T oc s~ 0 - 5 P i x e l d imens i on (w) <TNct ocew (orw-065) Slice thickness (h) <TNCT OC h~05 T a b l e 6.3: F r o m the results i n table 6.2, the percentage noise relat ive to the reference pro toco l (Re l . c r j v c r ) c a n be der ived for any i m a g i n g pro toco l . G i v e n a noise measurement (Meas. c r ^ C T ) for the reference protoco l , image noise can t h e n be ca l cu la ted for any C T i m a g i n g pro to co l (Ca l c . <TJV c t ) . E x a m p l e s for low, average a n d h i g h noise protocols are prov ided . I m a g i n g Non-ref . Meas . UNCT R e l . O~NCT C a l c . c r / v C T Diff. p ro toco l parameters (H) (%) (H) (H) Reference — 3.64 100 3.64 — L o w noise 3 0 0 m A , 1 0 m m , 4s, S O F T 0.87 19 0.69 0.18 Average noise 1 2 0 k V , 2 5 0 m A , 5 m m , 2s 2.53 68 2.49 0.04 H i g h noise 1 0 0 m A , 1 m m , 0.8s, D E T A I L 32.24 878 32.00 0.24 6.2.5 I m a g e U n i f o r m i t y T h e a b i l i t y of a gel dos imetry read-out technique to prov ide s p a t i a l l y u n i f o r m images w h e n i m a g i n g a s t ruc ture free ob ject is c r i t i c a l for accurate gel d o s i m e t r y as n o n -u n i f o r m i t y c o u l d be falsely in terpre ted as dose inhomogene i ty i n a n image of a n i r r a d i a t e d gel. F i g u r e 6.6a shows the m e a n a n d s t a n d a r d d e v i a t i o n for 36 s p a t i a l l y d i s t inc t 25 x 25 p i x e l R O I s i n a single, b a c k g r o u n d s u b t r a c t e d C T image of a water f i l led p h a n t o m . T h e reference C T s c a n n i n g p r o t o c o l (table 6.1) was used. It is clear t h a t the v a r i a t i o n between R O I s ( s t a n d a r d d e v i a t i o n of 0.14 H ) is far less t h a n the uncer ta in ty w i t h i n each R O I (average s t a n d a r d d e v i a t i o n of 3.65 H ) . I n order to ensure that th i s u n i f o r m i t y holds after a p p l i c a t i o n of image averaging to reduce image noise, the test was repeated after e m p l o y i n g 50 image averages. T h e results , shown i n figure 6.6b, ind i ca te t h a t even w i t h the low noise present i n a large number of averaged images, the u n i f o r m i t y remains better t h a n the i n t r a - R O I v a r i a t i o n as the s t a n d a r d d e v i a t i o n between the R O I means is o n l y 0.03 H . T h i s excellent u n i f o r m i t y makes x - r a y C T a n a t t rac t i ve a l t e rnat ive to M R I a n d O C T gel i m a g i n g techniques w h i c h experience higher levels of n o n - u n i f o r m i t y w h e n not accounted for [32, 33, 133]. 6.2.6 I m a g i n g T i m e Single slice i m a g i n g T h e t i m e requ i red to image a single slice t h r o u g h a gel dos imeter is s i m p l y the t i m e per slice (s) t imes the number of averages (NAX) used to o b t a i n the final image. I n contrast to vo lume i m a g i n g (below) factors such as x - r a y t u b e h e a t i n g w i l l not t y p i c a l l y restr ic t the technique used for s cann ing a single image. Consequent ly , C T scan technique s h o u l d be m a x i m i z e d (h igh mA, NAX) w h e n scanning a single slice t h r o u g h a gel. T h i s w i l l p rov ide a n image o p t i m i z e d for low noise a n d s t i l l a l l ow the image to be acqu ired i n a few minutes . T h i s technique c a n be used to m a x i m i z e dose reso lut ion i n a reg ion of p a r t i c u l a r interest i n the gel. rj o c a c (0 2 i 1 r SINGLE IMAGE 10 15 20 25 30 35 ROI (25x25 pixels) (a) 4 -2 --2 --4 -50 IMAGE AVERAGES 10 15 20 25 30 35 ROI (25x25 pixels) (b) F i g u r e 6.6: S p a t i a l u n i f o r m i t y i n C T images as measured u s i n g 36 s p a t i a l l y d i s t i n c t 2 5 x 2 5 p i x e l R O I s i n a single image (a) a n d 50 averaged images (b). B a c k g r o u n d s u b t r a c t i o n was app l i ed i n b o t h cases. E r r o r bars represent the s t a n d a r d d e v i a t i o n o n the m e a n for each R O I . V o l u m e i m a g i n g I m a g i n g t i m e for vo lume s c a n n i n g depends o n the n u m b e r of slices acqu ired a n d therefore o n s p a t i a l reso lut ion requirements a n d the l e n g t h of the p h a n t o m be ing imaged . However , for a t y p i c a l single slice C T scanner, the t o t a l t i m e requ i red to image a g iven l ength of gel is not equal to the single slice scan t i m e (as descr ibed above) m u l t i p l i e d b y the number of slices required . T h i s is due to t u b e over -heat ing a n d the resu l t ing coo l ing t i m e t h a t is necessary for m u l t i - i m a g e . s c a n n i n g . F o r a g iven system, the amount of t u b e heat ing a n d hence the t i m e requ i red for coo l ing depends o n the scan technique employed . A s a result , for vo lume i m a g i n g , scan technique is c r i t i c a l for d e t e r m i n i n g the t i m e requ i red to C T image a g iven v o l u m e of gel a n d , consequently, i m a g i n g t i m e involves a compromise w i t h image noise. T h e re lat ionships between t u b e l o a d , i m a g i n g t i m e a n d image noise are invest igated below. T h e l o a d on a n x - r a y t u b e is g iven b y kJ = kV x mA x s x NAX, showing t h a t tube voltage, current , slice scan t i m e a n d the n u m b e r of image averages a l l affect tube l oad equally . G i v e n the results of sect ion 6.2.2, m a x i m i z i n g kV is a n o p t i m u m strategy for efficient noise reduc t i on , but the r e m a i n i n g technique factors (mA, s, NAX) must be chosen based o n a compromise w i t h i m a g i n g t ime . F i g u r e 6.7 shows the e x p e r i m e n t a l re la t i onsh ip between t u b e l o a d , ca l cu la ted based o n scan technique, a n d the e x p e r i m e n t a l l y de te rmined t i m e required to image each slice w h e n sequential ly s cann ing m u l t i p l e slices. T h e t imes shown i n figure 6.7 are stable , per slice i m a g i n g t imes that are reached once the t u b e has reached i ts m a x i m u m opera t ing temperature . I n each case, the t i m e requ i red to image the first ~ 3 slices was a b i t less as the t u b e was s t i l l w a r m i n g u p . T h e measurements were o b t a i n e d b y v a r y i n g mA, s a n d NAX wh i l e us ing an otherwise i d e n t i c a l s cann ing pro toco l , as descr ibed i n sect ion 6.1.1. T h e l inear result i n figure 6.7 indicates t h a t for a n x - r a y t u b e o p e r a t i n g at i ts m a x i m u m temperature the t i m e required to image sequent ia l slices is d i r e c t l y I • I • I 1 I 1 I I I I I . I 200 300 400 500 600 700 800 900 X-ray tube load (kJ) F i g u r e 6.7: T h e average t i m e required to o b t a i n a single slice image w h e n sequen-t i a l l y s canning m u l t i p l e slices us ing a single slice C T scanner. T h e t i m e depends on the l oad p laced on the x - r a y t u b e a n d therefore on C T scan technique. T h i s is due to requ ired tube coo l ing between slices. p r o p o r t i o n a l to the l oad on the t u b e a n d hence the scan technique used for i m a g i n g . T h i s s imple re la t ionship means t h a t t o t a l i m a g i n g t i m e can be easi ly pred i c ted for a g iven i m a g i n g protoco l . C o m b i n i n g th is result w i t h the effect of scan pro toco l on image noise (section 6.2.2), the e x p e c t a t i o n is tha t , for a g iven scan k V , image noise is a p p r o x i m a t e l y re lated to i m a g i n g t i m e (time) by: C T J V c t oc 1/s/time. T h i s is a n i m p o r t a n t re la t ionship since i t provides a means of choos ing a n o p t i m u m C T s cann ing pro to co l i n cases where i m a g i n g t i m e constra ints prevent m a x i m i z i n g technique for noise reduc t i on . F o r example , f r o m figure 6.7, g iven a 1 hour t i m e constra int t o scan 20 slice pos i t ions w i t h i n a gel , a scan technique p r o v i d i n g a tube l o a d < 600 k J must be used. U s i n g the above re la t i onsh ip , th i s constra int w i l l force a n a p p r o x i m a t e d o u b l i n g of image noise c o m p a r e d to the non - res t r i c t i ve , m a x i m u m technique s canning that is possible w h e n p e r f o r m i n g single slice i m a g i n g . M o s t cancer hospi ta ls are cur rent ly equ ipped w i t h C T s imula tors for t rea t -ment p l a n n i n g w h i c h are single slice machines s i m i l a r to the scanner used i n these studies. A s such, the need to c o n s t r a i n i m a g i n g technique i n a compromise w i t h i m a g i n g t i m e , as descr ibed above, w i l l be encountered b y most centres w h e n i m p l e -m e n t i n g C T gel dosimetry . However , the future of C T s i m u l a t i o n sure ly includes m u l t i - s l i c e scanning . M o d e r n mul t i - s l i c e d iagnost i c C T scanners can t y p i c a l l y scan 16 or more slices s imultaneously . C l e a r l y , the use of such a sys tem for C T gel dos ime-t r y w o u l d s igni f i cant ly reduce the t o t a l i m a g i n g t i m e requ i red for vo lume scanning . T h i s is a avenue for future work , as descr ibed i n chapter 8. 6.2.7 A c h i e v a b l e D o s e R e s o l u t i o n T h e results of the studies descr ibed i n th i s w o r k are c o m b i n e d here to ca lculate the dose reso lut ion current ly achievable for different P A G C T d o s i m e t r y systems, g iven voxe l size a n d i m a g i n g t i m e constra ints . Dose reso lut i on is ca l cu la ted , as descr ibed i n sect ion 3.5, us ing image noise levels ( C J V c t ) es tabl ished i n this s t u d y a n d C T dose responses establ ished for various P A G compos i t i ons i n chapter 5. Tab le 6.4: A compar i son of re lat ive dose reso lut ions for a range of P A G dosimeter f o rmulat ions used i n the i r l inear or "quas i - l inear " regions. A representat ive value of image noise, 0.3 H , was used i n a l l cases. P A G f o r m u l a t i o n slope (HGy-1) £>A,% (%) « (%) 0 % C , 6 % T 20 0.23 6.6 18.4 5 0 % C , 6 % T 8 0.83 4.5 12.5 7 0 % C , 6 % T 20 0.24 6.2 17.2 1 0 0 % C , 3 % T 100 0.04 7.7 21.3 A c ompar i son of achievable dose reso lut ions for the P A G formulat ions i n -vest igated i n sect ion 5.3 is p r o v i d e d i n tab le 6.4. A value of 0.3 H for ONCT is used for a l l ca l cu lat ions i n order to d i r e c t l y compare the re lat ive per formance of the dosimeters . T h i s value is chosen as representat ive image noise such as m a y be achieved for a t y p i c a l gel p h a n t o m us ing a n o p t i m i z e d h i g h technique pro to co l a n d image averaging. However , as discussed i n th is work , th i s va lue c a n v a r y great ly as i t is dependent o n m a n y factors. T h e l inear dose ranges were selected to meet a m i n i m u m goodness of fit requirement : B? = 0.98. A s s h o w n i n tab le 6.4, the 50 % C P A G provides the best dose reso lut ion . T h i s is i n spite of th i s gel h a v i n g the smallest l inear dose range. I n contrast , the 100 % C P A G , despite e x h i b i t i n g the longest l inear dose range, has the poorest dose reso lut ion . However , the l ong l inear range of th is gel remains interes t ing a n d i t m a y f i n d use i n low reso lut i on eva luat i on of very h i g h dose a n d dose range app l i ca t i ons , such as brachytherapy . T h e s t u d y of P A G dens i ty to dose response, presented i n sect ion 5.4, i l l u m i -n a t e d a l inear increase i n the dose response of P A G s of a g iven % C w i t h increas ing % T [144]. T h i s , i n theory, c o u l d help improve the P A G dose resolut ions as l i s t ed i n tab le 6.4. U n f o r t u n a t e l y , the re lat ive i n s o l u b i l i t y of b is w i l l l i m i t increas ing % T i n m a n y of these f ormulat ions . Some authors have r e p o r t e d m a k i n g 50 % C P A G s w i t h h i g h % T [43], b u t these gels are dif f icult a n d i m p r a c t i c a l to make , often r e q u i r i n g s t i r r i n g overnight . U s i n g the best per f o rming P A G f o r m u l a t i o n as d e t e r m i n e d f r o m the above T a b l e 6.5: A compar i son of re la t ive dose resolut ions for C T P A G d o s i m e t r y con-s t ra ined to meet a range of voxe l sizes. These results are not c ons t ra ined b y i m a g i n g t ime . A 6 % T , 50 % C P A G was used. V o x e l d i m e n s i o n x , y ( m m ) z ( m m ) DAy0 (%) (%) 0.5 1.0 11.5 31.7 1.0 1.0 9.2 25.4 0.5 3.0 6.8 18.8 1.0 3.0 5.4 15.0 1.5 3.0 4.2 11.5 2.0 3.0 3.0 8.2 2.5 3.0 2.2 6.2 2.5 5.0 1.8 5.1 5.0 5.0 1.0 2.8 5.0 10.0 0.7 1.9 results (50 % C , 6 % T ) , tab le 6.5 summar izes the achievable dose reso lut i on at var ious voxe l sizes for th is system. These results are not constra ined b y t i m e a n d are therefore appl i cab le to i m a g i n g select regions of interest w i t h i n the gel as opposed to complete vo lume i m a g i n g . T h e compromise between u s i n g a s m a l l voxe l size a n d i m p r o v i n g dose reso lut ion is c lear. T h e improvement i n £>A,% w h e n voxe l size is l i m i t e d to tha t of m a n y t reatment p l a n n i n g dose ca l cu lat ions (2.5 x 2.5 x 3 m m 3 ) is h i g h l y signif icant: f r o m 11.5 % to 2.2 %. Consequent ly , for ver i f i cat ion of such t reatment p lans , th is type of voxe l size r e d u c t i o n s trategy is ent i re ly appropr ia te . I n general , the results i n tab le 6.5 ind i ca te the a b i l i t y of C T P A G d o s i m e t r y to achieve excellent dose resolut ions i n c r i t i c a l regions of interest i n a dose vo lume . I n a d d i t i o n , a l t h o u g h a r educ t i on i n voxe l size is c u r r e n t l y requ ired i n order to achieve c l i n i c a l l y acceptable dose reso lut ions , these results i l lus t ra te the p o t e n t i a l for C T gel dos imetry to produce 3 D dose measurements o n a fine (< 1 m m ) voxe l g r i d . T a b l e 6.6 shows the dose reso lu t i on achieved for th is same gel , over the same range of voxel sizes as table 6.5, but w i t h the a d d e d requirement of i m a g i n g a n entire gel vo lume w i t h i n 1 hour . A I L (10 x 10 x 10 c m 3 ) gel vo lume is used T a b l e 6.6: A compar i son of re la t ive dose resolut ions for C T P A G d o s i m e t r y con-s t ra ined to meet a range of voxe l sizes for i m a g i n g a 1 L vo lume of gel (10 x 10 x 10 c m 3 ) w i t h i n a 1 hour t i m e l i m i t . A 6 % T , 50 % C P A G was used. V o x e l d imens ion R e q u i r e d n u m b e r x , y (mm) z (mm) of images DA,% (%) D$* (%) 0.5 1.0 100 37.7 104.3 1.0 1.0 100 30.1 83.5 0.5 3.0 33 14.9 41.2 1.0 3.0 33 11.9 33.0 1.5 3.0 33 9.1 25.3 2.0 3.0 33 6.5 18.0 2.5 3.0 33 4.9 13.6 2.5 5.0 20 3.6 10.0 5.0 5.0 20 2.0 5.5 5.0 10.0 10 1.0 2.8 to represent a t y p i c a l , c l i n i c a l l y re levant vo lume . A s discussed i n sect ion 6.2.6, a n i m a g i n g t i m e constra int for vo lume i m a g i n g requires l i m i t i n g C T i m a g i n g technique w h i c h produces increased image noise a n d , as g iven b y equat i on 3.25, decreased dose reso lut ion . T h i s po int is c l ear ly i l l u s t r a t e d i n tab le 6.6 where i m a g i n g w i t h a s m a l l voxe l size provides d r a m a t i c a l l y reduced dose reso lut ion c o m p a r e d to us ing a larger voxe l size. T h e effect is so large due to the c o m b i n a t i o n of two effects. T h e first effect derives d i r e c t l y f r o m r e d u c t i o n i n voxe l size (as descr ibed i n sect ion 6.2.3). T h e second is due to the increased n u m b e r of images requ i red to image a l ength of gel w i t h a s m a l l slice spac ing . Pos i t i ve ly , the results ind i ca te t h a t even w i t h i n a s tr i c t 1 hour t i m e l i m i t , < 5% c a n be achieved u s i n g a voxe l size of 2.5 x 2.5 x 3.0 m m 3 , w h i c h is consistent w i t h the c a l c u l a t i o n g r i d size for m a n y t reatment p l a n n i n g dose ca l cu lat ions . It is evident t h o u g h that vo lume i m a g i n g w i t h a s m a l l voxel size ( p a r t i c u l a r l y w i t h a slice separat ion of 1 m m ) cannot c u r r e n t l y prov ide c l i n i c a l l y useful results w i t h i n one hour . However , the future i n t e g r a t i o n of mul t i - s l i c e C T scanners into cancer care w i l l s t a n d to s ign i f i cant ly improve th is result . 6.2.8 C o m p a r i s o n w i t h M R I a n d O C T G e l D o s i m e t r y T a b l e 6.7 l ists a c o m p a r i s o n of current ( typical ) dose reso lut i on a n d s cann ing t i m e capabi l i t i es of M R I , O C T , a n d x - r a y C T gel dosimetry . It m u s t be n o t e d t h a t th i s is a rough c ompar i son as due to the m a n y factors tha t can affect results for a l l i m a g i n g moda l i t i e s , a direct c o m p a r i s o n of o p t i m i z e d systems is d i f f i cult . T h e a i m here is to give the best possible es t imate of the capabi l i t i es of C T as c o m p a r e d w i t h M R I a n d O C T . V o x e l sizes are q u o t e d to fac i l i ta te c ompar i son since, as s h o w n here for x - r a y C T , for a l l moda l i t i es dose reso lu t i on a n d i m a g i n g t i m e w i l l d e p e n d on voxe l size. U s i n g M R I , as g iven i n tab le 6.7, G u s t a v s s o n et al. (2004) are able to achieve a m i n i m u m dose reso lut ion of ~ 4 % (95% confidence). T h i s is a n impress ive result w h e n c o m p a r e d to 6 7 % confidence i n t e r v a l values quoted i n tab le 6.7 [112]. However , th is value is a m i n i m u m a n d increases for different doses. A l s o , the s can thickness for th i s measurement was set to 5 m m . O l d h a m et al. (2001) quote a dose reso lut i on of ~ 4 . 5 % for the ir M R I measurement [14]. I n a theoret i ca l work , DeDeene a n d B a l d o c k (2002) show that a dose reso lut i on of 3.6% (95% confidence) s h o u l d be achievable u s i n g a n o p t i m i z e d M R I pulse sequence. However , slice th ickness is not i n d i c a t e d i n th i s work [120]. I n t e rms of O C T , O l d h a m et al. (2001) quote a m i n i m u m dose reso lu t i on of 1.1% w i t h 1 m m 3 voxels. T o achieve th i s reso lut i on , however, r equ i red 20 m i n of d a t a acqu i s i t i on per image slice, m a k i n g acquis i t ions of 3 D dose vo lumes i m p r a c t i c a l [14]. X u et al. (2003) quote a dose reso lut i on of 4%, u s i n g 2 x 2 x 2 m m 3 voxels a n d 5-6 m i n u t e slice a c q u i s i t i o n t imes [128]. I s l a m et al. (2003) quote a dose reso lut ion of 3 % us ing 1.4 x 1.4 x 1 m m 3 voxels a n d 12 m i n u t e acquis i t ions per image slice [129]. X - r a y C T dose reso lut i on is s t i l l not at the leve l of e ither M R I or O C T . I n this work , dose reso lu t i on is 5% us ing 1 x 1 x 3 m m 3 voxels a n d th i s value increases to 15% for a 9 5 % confidence i n t e r v a l measurement . However , i t must be no ted that the a c q u i s i t i o n t i m e r e q u i r e d for a single slice is less t h a n 1 m i n u t e , m a k i n g acqu is i t i on t i m e s igni f i cant ly faster t h a n e i ther M R I or O C T . T h i s makes the r a p i d acqu is i t i on of v o l u m e t r i c d a t a most feasible w i t h x - r a y C T . A s descr ibed T a b l e 6.7: A compar i son of the dose reso lut ion a n d t y p i c a l s can t imes for M R I , O C T , a n d x - r a y C T gel dos imetry . V o x e l sizes are g iven to fac i l i ta te compar ison . A l l quo ted results are for P A G dosimeters . M o d a l i t y V o x e l size A p p r o x dose T i m e Ref . ( m m ) resol . (%)° (min / s l i c e ) M R I 1x1x2 4 .5% — O l d h a m et al. 2001 1x1x5 4%e,f 25 G u s t a v s s o n et al. 2004 — 3.6% e '3 34 D e D e e n e a n d B a l d o c k 2002 O C T l x l x l 6 1.1% 20 O l d h a m et al. 2001 2 x 2 x 2 c 4% 5-6 X u et al. 2003 1 . 4 x l . 4 x l d 3% 12 I s l a m et al. 2003 X - r a y C T 1x1x3 5% <1 this w o r k a based on re lat ive dos imetry . 6 7 % confidence i n t e r v a l q u o t e d unless otherwise noted . b thickness based o n laser spot size (~0.75 m m ) c var iab le d o w n to 0.3 m m d 0.8 m m possible (laser spot size) e 9 5 % confidence f varies w i t h dose, m i n i m u m quoted 9 theoret i ca l result above, improvement i n the dose reso lut i on can be made , i n p a r t , b y longer scan t imes , a l t h o u g h the p r a c t i c a l l i m i t a t i o n s of t u b e h e a t i n g must be respected w h e n p e r f o r m i n g 3 D dosimetry . 6.3 Chapter Summary T h e studies presented i n th i s chapter have developed o p t i m i z a t i o n strategies for noise r e d u c t i o n i n C T p o l y m e r gel dos imetry as w e l l as assessed sys tem perfor-mance capabi l i t ies . P h a n t o m design, C T i m a g i n g technique a n d voxe l d imensions were a l l demonstra ted to be i m p o r t a n t considerat ions for noise r educ t i on . I n p a r -t i c u l a r , s m a l l p h a n t o m sizes are r e commended wherever poss ib le , increas ing x - r a y t u b e voltage shou ld be p r i o r i t i z e d for low noise scan technique a n d larger voxe l sizes (for example 2 x 2 x 3 m m 3 ) shou ld be considered i n some s i tuat ions . T h e re la t i onsh ip between i m a g i n g t i m e a n d image noise was assessed a n d real ist ic noise l i m i t s demonst ra ted for vo lume scanning . U s i n g these results , achievable dose res-o l u t i o n was ca l cu la ted for the P A G formulat ions s t u d i e d i n chapter 5 g iven voxe l size a n d i m a g i n g t i m e . I n s u m m a r y , i n d e t e r m i n i n g o p t i m i z a t i o n strategies for noise reduc t i on , the w o r k i n th i s chapter has made a s ignif icant step towards i m p r o v i n g the dose reso lut ion achievable w i t h C T gel dos imetry . P u t in to context w i t h other gel dos imetry read-out moda l i t i e s ( M R I a n d O C T ) , C T read-out is shown to be c o m p e t i t i v e l y fast but to s t i l l l a g b e h i n d i n dose reso lut ion . However , as w i l l be shown i n the fo l lowing chapter , d i g i t a l image f i l t er ing can be used to further i m -prove dose reso lut i on a n d future work i n gel development (chapter 8) promises to produce a d d i t i o n a l improvements . Chapter 7 Digital Image Filtering Studies T h i s chapter presents studies pre formed to invest igate the p o t e n t i a l of d i g i t a l image f i l t e r ing for r educ ing noise i n C T images of i r r a d i a t e d p o l y m e r gel . T h i s w o r k has two components . F i r s t , C T noise as present i n gel images is charac ter i zed a n d th is i n f o r m a t i o n is used to select appropr ia te f i lters for inves t igat i on . Second, the selected filters are tested on a C T image of a gel i r r a d i a t e d w i t h a stereotact ic rad iosurgery ( S R S ) dose d i s t r i b u t i o n . T h i s image is f r o m a prev ious s t u d y [41, 96] a n d was selected as i t is shows a c l i n i c a l l y relevant dose d i s t r i b u t i o n a n d d isp lays a low contrast to noise level t y p i c a l of C T gel dos imetry . F i l t e r per formance is eva luated based on: (1) the achieved improvement i n dose reso lu t i on a n d (2) the a b i l i t y to m a i n t a i n the spa t ia l in tegr i ty of the S R S dose d i s t r i b u t i o n . E x p e r i m e n t a l detai ls of the noise charac ter i zat i on , d i g i t a l image filtering a n d image analys i s as we l l as the b a c k g r o u n d gel w o r k are p r o v i d e d i n sect ion 7.1. R e s u l t s of the noise charac te r i za t i on a n d filter per formance analys is are p r o v i d e d i n sections 7.2 a n d 7.3, respectively. A deta i led descr ip t ion of the filters invest igated is p r o v i d e d i n a p p e n d i x A . T h e w o r k presented i n th is chapter resu l ted i n a research paper p u b l i s h e d i n the i n t e r n a t i o n a l refereed j o u r n a l Medical Physics [143]. 7.1 Experimental Details 7.1.1 B a c k g r o u n d G e l W o r k These studies employ the C T images of two i r r a d i a t e d P A G s f r o m a prev ious w o r k [41, 96] . T h e first is the image of a gel i r r a d i a t e d w i t h a c l i n i c a l t reatment ( S R S ) a n d the second is the image of a c a l i b r a t i o n gel used to convert the S R S image to a dose m a p . T h e effectiveness of d i g i t a l f i l t er ing was tested b y a p p l y i n g each f i l ter to the S R S gel image. T h i s image was selected since, u p o n convers ion to dose, i t p r o v i d e d a relevant c l i n i c a l dose d i s t r i b u t i o n a n d , w i t h a contrast t o noise level representat ive of C T gel dosimetry , p r o v i d e d a rea l i s t i c s i t u a t i o n i n w h i c h to test filter per formance . B o t h gels were made in-house f r o m a single b a t c h of P A G , c o m p o s i t i o n 6 % T , 50 % C . T h e c l i n i c a l gel was i r r a d i a t e d w i t h a four arc S R S t rea tment to a m a x i m u m dose of 15 G y a n d the c a l i b r a t i o n gel w i t h 10 regions of u n i f o r m dose ( f rom 1 to 24 G y ) . B o t h P A G s were i m a g e d us ing a H i S p e e d C T / i C T scanner ( G E M e d i c a l Systems) , descr ibed i n sect ion 4.3.1, a n d the fo l l owing pro toco l : 120 k V , 200 m A s , 3 m m slice spacing, 40 c m field of v i e w a n d d e t a i l r e construc t i on . 16 images were averaged to improve contrast to noise levels a n d a b a c k g r o u n d s u b t r a c t i o n was per fo rmed to remove b e a m h a r d e n i n g a n d r i n g art i facts . T h e final processed S R S a n d c a l i b r a t i o n gel images, i n a d d i t i o n to a single, unprocessed S R S image (used for noise charac ter i za t i on i n sect ion 7.1.2) are shown i n figure 7.1. T h e c l i n i c a l image was converted to a dose m a p u s i n g a dose response curve ex t rac ted f r om the c a l i b r a t i o n gel. T h e c a l i b r a t i o n curve is shown i n figure 7.2. T h e f u l l response is mono -exponent ia l , however i t c a n be d i v i d e d in to two l inear regions defined b y a n R 2 > 0.98: a low dose region ( L D R ) , < 12 G y , a n d a h i g h dose region ( H D R ) > 12 G y . T h e L D R is h i g h l y l inear ( R 2 = 0.995), w i t h a slope of 85 ± 1 c G y / H . T h e H D R is also very l inear ( R 2 = 0.985), w i t h near ly twice the slope of the L D R , 164 ± 14 c G y / H . T h e dose reso lut i on results presented i n sect ion 7.3.1 are ca l cu la ted f r om noise measurements a n d u s i n g these l inear slopes, as descr ibed (a) (b) (c) F i g u r e 7.1: C T images of the S R S gel a n d the c a l i b r a t i o n gel: a) a single, u n p r o -cessed S R S gel image, b) the processed (averaged a n d b a c k g r o u n d subt rac ted b u t unfi ltered) S R S gel image a n d c) the s i m i l a r l y processed c a l i b r a t i o n gel image. i n sect ion 3.5. 7.1.2 N o i s e C h a r a c t e r i z a t i o n A s was discussed i n 2.4.1, C T noise is a c o m b i n a t i o n of q u a n t u m noise, e lectronic noise a n d noise due to the image generat ion process a n d , as demonst ra ted i n chapter 6, depends on m a n y factors ( imag ing pro toco l , r e cons t ruc t i on a l g o r i t h m etc.) . C o n -sequently, the noise i n C T images is t y p i c a l l y charac ter i zed t h r o u g h measurement . O n e s t a n d a r d technique used to character ize image noise is to o b t a i n a h i s t o g r a m of C T numbers (NCT) f r o m a u n i f o r m (no contrast ) area of a n image. T h e result is ca l led a noise p r o b a b i l i t y dens i ty f u n c t i o n ( P D F ) a n d analys is of its f o r m c a n i l l u m i n a t e the type of noise degrading the image: e.g. U n i f o r m , Po i s son , G a u s s i a n etc. [206] and , for the purposes of th is work , fac i l i tate select ion of su i table f i lters. A s descr ibed below, i n th is w o r k P D F s were e x t r a c t e d f r om images of a c o m m e r c i a l C T p h a n t o m a n d the S R S gel images s h o w n i n figure 7.1. C T P h a n t o m A c o m m e r c i a l , so l id water C T q u a l i t y assurance p h a n t o m ( R M I H e a d / B o d y P h a n -t o m M o d e l 4 6 1 A , G A M M E X - r m i , M i d d l e t o n W I ) was used to character ize C T image 0 5 10 15 20 25 NCT (H) F i g u r e 7.2: Dose to Ncr c a l i b r a t i o n curve o b t a i n e d f r o m the c a l i b r a t i o n gel shown i n figure 7.1c. NCT is the value above b a c k g r o u n d . noise for the scanner a n d i m a g i n g p r o t o c o l used to image the S R S gel (see sect ion 7.1.1). T h e p h a n t o m was scanned 6 t imes , each t i m e w i t h a n insert of different but u n i f o r m dens i ty m a t e r i a l p laced at i t s centre. T h e purpose of i m a g i n g a var ie ty of mater ia l s was to determine i f image noise level var i ed w i t h m e a n NCT (a charac -ter is t i c of P o i s s i o n d i s t r i b u t e d noise) or was re la t i ve ly constant across a l l mater ia l s imaged . T h e mater ia ls used were po lyethy lene , po lypropy lene , water , a c ry l i c , tef lon a n d bone, p r o v i d i n g a n NCT range f r o m ~ -100 to 1200 H . F i g u r e 7.3 shows a n example image of the C T p h a n t o m w i t h the a c r y l i c insert . F o r each insert , us ing M a t L a b , a P D F was ca l cu la ted f r o m a 256 p i x e l reg ion of interest at the centre of the insert . T h e P D F s were e x p o r t e d to , a n d fit us ing , O r i g i n 7 . 0 . S R S G e l Images Noise character ist ics were also measured for the o r i g i n a l a n d processed S R S gel images (as shown i n figures 7.1a a n d b) . P D F s of b a c k g r o u n d noise were ob ta ined F i g u r e 7.3: C T image of the so l id water C T p h a n t o m used to character ize C T image noise. T h i s example inc ludes the a c r y l i c insert . f r om the same 256 p i x e l reg ion of interest for b o t h images us ing M a t L a b a n d fit us ing O r i g i n 7 . 0 . Noise c o u l d not be accurate ly measured i n the h i g h dose region of the gel due to dose heterogeneity a n d therefore image n o n - u n i f o r m i t y i n this region. 7.1.3 D i g i t a l I m a g e F i l t e r i n g D i g i t a l image f i l ter ing can be per f o rmed i n either the frequency or s p a t i a l domains . T h e results of the noise charac te r i za t i on , discussed i n sect ion 7.2, i n d i c a t e d that spa t ia l f i l ter ing techniques for r e m o v i n g s p a t i a l l y invar iant , add i t i ve G a u s s i a n noise shou ld be most effective for C T gel d o s i m e t r y a n d th is type of f i l t er ing is evaluated i n th is work [206]. T h e r e are m a n y such fi lters t h a t , to v a r y i n g degrees, reduce noise a n d preserve edges a n d i t is b e y o n d the scope of th is thesis to evaluate a l l p o t e n t i a l f i lters. C lass i c noise r e d u c t i o n a n d edge preservat ion fi lters are eva luated alongside some complex filters t h a t , based o n a l i t e ra ture review, are expected to per form wel l [207, 208]. I n a d d i t i o n , a l l selected filters are c o m p u t a t i o n a l l y efficient. T h e eva lu -a t e d filters are: m e a n ( M E A N ) , m e d i a n ( M E D I A N ) , m i d p o i n t ( M I D P O I N T ) , m e a n adapt ive ( A D A P T I V E ) , a l p h a - t r i m m e d m e a n ( a - M E A N ) , s i g m a m e a n ( S I G M A ) a n d a re la t ive ly new filter ca l l ed S U S A N . D e t a i l e d descr ipt ions of these filters are p r o v i d e d i n a p p e n d i x A . E a c h filter was tested o n the processed S R S gel image (figure 7.1b). T h e effect of mask size was e x a m i n e d by u s i n g 3 x 3, 5 x 5 a n d 7 x 7 p i x e l mask sizes. T h e o n l y except ion was the S U S A N filter w h i c h , since i ts per formance is independent of mask size [207], was a p p l i e d for a 3 x 3 p i x e l mask only. A l l filtering was per formed us ing M a t L a b . T h e M E A N , M E D I A N a n d A D A P T I V E filters are avai lable as b u i l t - i n M a t L a b funct ions a n d the r e m a i n i n g f i lters were i m p l e m e n t e d u s i n g user-defined code. Theore t i ca l ly , the M E A N fi lter is expected to prov ide the best noise r educ t i on but a n unacceptab le level of degradat i on of s p a t i a l dose i n f o r m a t i o n . T h e other filters invest igated are a l l expected to better m a i n t a i n the s p a t i a l in tegr i ty of the dose at least to some degree. G i v e n i ts c l a i m to c ombine the best of the character ist ics of m o d e r n edge preserv ing filters, the S U S A N filter is hypothes ized to prov ide the best overal l results . 7.1.4 I m a g e A n a l y s i s T h e f i l tered images were analysed q u a n t i t a t i v e l y to assess the achieved dose reso lu -t i o n a n d the effect o n the s p a t i a l d i s t r i b u t i o n of dose. D a t a e x t r a c t i o n a n d analys is was per formed us ing M a t L a b a n d i n some cases O r i g i n 7 . 0 . T h e same, u n i f o r m , 256 p i x e l reg ion of interest i n the b a c k g r o u n d reg ion of each image was used to ex t rac t image noise, <TNCT- F r ° m these (JNCT values, <7£> a n d were ca l cu la ted u s i n g equat ions 3.25 a n d 3.28 a n d the dose c a l i b r a t i o n curve shown i n figure 7.2. T h e ef-fect of the filters o n the s p a t i a l d i s t r i b u t i o n of dose was eva luated us ing (1) profi les t h r o u g h the H D R a n d (2) dose area h i s tograms ( D A H s ) . A D A H records the areas covered b y g iven doses. F r o m the profi les, the w i d t h s of the H D R at 750 c G y (1 /2 the m a x i m u m dose) a n d of the r i g h t - h a n d H D R p e n u m b r a ( from 1000 to 500 c G y ) were ca l cu lated . F r o m the D A H s , dose areas ( A ^ ) , the area t h a t received greater t h a n or equal to a g iven percentage dose ( D ) , were der ived for 100, 80 a n d 5 0 % of 1500 c G y . A l l results were ca l cu la ted a n d c o m p a r e d for the f i l tered images. F i g u r e 7.4: P D F s for 6 u n i f o r m mater ia l s w i t h v a r y i n g C T contrast : a) p o l y p r o p y -lene, b) po lyethylene , c) s o l id water , d) a c ry l i c , e) tef lon a n d f) bone. A l l P D F s are fit w i t h Gauss ians . Resu l t s ind i ca te that image noise is r e la t ive ly independent of s igna l s t rength . 9 Averaged and Background Subtracted Image b j - Gaussian Fit CT Number (H) F i g u r e 7.5: P D F s for a b a c k g r o u n d region of a) the single, unprocessed S R S gel image a n d b) the averaged a n d b a c k g r o u n d s u b t r a c t e d S R S gel image (shown i n figure 7.1a a n d b respect ive ly ) . B o t h P D F s are fit w i t h G a u s s i a n d i s t r i b u t i o n s . N o t e that i n b) NCT is g iven re lat ive to the b a c k g r o u n d . 7.2 Results and Discussion I: Characterization of Image Noise 7.2.1 C T P h a n t o m P D F s for the s ix C T p h a n t o m inserts are shown i n figure 7.4. O n v i s u a l inspec t i on the P D F s appear to be n o r m a l l y d i s t r i b u t e d a n d u p o n f i t t i n g , are wel l descr ibed by Gauss ians . Over a l l mater ia l s tested, the m e a n NCT var i ed b y ~ 1300 H (from -99 to +1202 H ) but the s t a n d a r d dev ia t i on o n these means var i ed by on ly 9.1 H ( from 15.9 to 25.0 H ) . A s a result , w h e n u s i n g the scanner a n d i m a g i n g protoco l def ined i n sect ion 7.1.2, C T image noise s h o u l d vary less t h a n 0.01 H per 1 H change i n image NCT (or by ~ 1 %) . F o r the s m a l l change i n NCT observed i n P A G ( m a x i m u m 20 H ) , i t is reasonable to consider the noise i n P A G C T images to be independent of NCT-7.2.2 G e l D o s i m e t e r T h e G a u s s i a n nature of the C T noise is con f i rmed for the S R S P A G images. F i g u r e 7.5 shows the P D F s ob ta ined for the b a c k g r o u n d region of a) the single, unprocessed S R S gel image a n d b) the averaged a n d b a c k g r o u n d s u b t r a c t e d S R S image (shown i n figures 7.1a a n d b, respect ive ly ) . B o t h P D F s are c lear ly w e l l m o d e l l e d b y G a u s s i a n d i s t r i b u t i o n s . D u e to n o n - u n i f o r m dose, noise i n the H D R c o u l d not be s i m i l a r l y measured . However , as discussed above (sect ion 7.2.1), noise i n th i s h igher contrast reg ion w o u l d be ind i s t ingu ishab le f r o m the noise i n the b a c k g r o u n d region. T h e consequence is tha t even t h o u g h C T images of i r r a d i a t e d P A G w i l l have v a r y i n g levels of image contrast throughout (due to dose heterogeneity) , noise levels s h o u l d r e m a i n u n i f o r m across the image. A s such, i n w h a t fol lows, f i lters for r e m o v i n g s p a t i a l l y invar iant noise are invest igated for noise r e d u c t i o n i n C T P A G images. 7.3 Results and Discussion II: Filter Performance T h e results of f i l ter ing the S R S gel image w i t h the selected fi lters (descr ibed i n a p p e n d i x A ) are shown i n figures 7.6 t h r o u g h 7.12. Q u a l i t a t i v e l y , the degree of s m o o t h i n g increases w i t h m a s k size a n d the different f i lters appear to reduce image noise a n d degrade s p a t i a l dose i n f o r m a t i o n to v a r y i n g degrees. These f i l tered images are q u a n t i t a t i v e l y c o m p a r e d i n sections 7.3.1 a n d 7.3.2 below. (a) (b) (c) F i g u r e 7.6: R e s u l t of f i l t er ing w i t h the m e a n fi lter ( M E A N ) u s i n g a 3 x 3, 5 x 5 a n d 7 x 7 p i x e l mask i n a) , b) a n d c) respect ively . (a) (b) (c) F i g u r e 7.7: R e s u l t of f i l t er ing w i t h the m e d i a n filter ( M E D I A N ) us ing a 3 x 3, 5 x 5 a n d 7 x 7 p i x e l mask i n a) , b) a n d c) respect ively . (a) (b) (c) F i g u r e 7.9: R e s u l t of f i l t e r ing w i t h the m e a n adapt ive f i l ter ( A D A P T I V E ) u s i n g a 3 x 3, 5 x 5 a n d 7 x 7 p i x e l mask i n a) , b) a n d c) respectively. (a) (b) (c) F i g u r e 7.10: R e s u l t of f i l t er ing w i t h the a - t r i m m e d m e a n f i l ter ( a - M E A N ) us ing a 3 x 3, 5 x 5 a n d 7 x 7 p i x e l m a s k i n a), b) a n d c) respectively. 7.3.1 D o s e R e s o l u t i o n M e a s u r e d uNCT a n o - ca l cu la ted values of au a n d are p r o v i d e d i n tab le 7.1 for a l l f i l tered images a n d the unf i l tered image. T h e values of or> a n d are quoted for the L D R a n d H D R as defined above, i n sect ion 7.1.1. A s expected , ( from equations 3.25 a n d 3.28), for a g iven measured ONCT b o t h ur j a n d D9^% are worse at h i g h t h a n at l ow doses. F o r a l l f i l ters, dose reso lut ion is d r a m a t i c a l l y i m p r o v e d compared to the unf i l t ered image. T h i s improvement is however s m a l l for the S I G M A filter c o m p a r e d to the other f i l ters. In a d d i t i o n , w i t h the except ion of the S U S A N fi lter, th i s improvement increases w i t h mask size. I n general , th i s effect is expected since a larger m a s k provides a greater s m o o t h i n g area. T h e best dose reso lut ion for b o t h the 3 x 3 a n d 5 x 5 masks sizes is achieved us ing the M E A N fi lter F i g u r e 7.12: R e s u l t of f i l ter ing us ing the smallest un iva lue segment a s s i m i l a t i n g nucleus a p p r o a c h to noise r e d u c t i o n ( S U S A N ) . S U S A N fi lter per formance is inde -pendent of mask size. w h i l e the best dose reso lut ion us ing a 7 x 7 m a s k is achieved us ing the M E D I A N f i l ter . T h e a - M E A N a n d A D A P T I V E fi lters also prov ide excellent improvement to dose reso lut ion , comparab le to, but s l i gh t ly poorer t h a n , the M E A N fi lter for a l l m a s k sizes. T h e S U S A N fi lter provides dose reso lut ions comparab le to the the 3x3 m a s k results of the a - M E A N a n d A D A P T I V E n i ters , b u t is outper f o rmed i f a larger m a s k is used. T h e M I D P O I N T fi lter per forms p o o r l y for a l l mask sizes. However , over a l l mask sizes the worst fi lter for i m p r o v i n g dose reso lu t i on is the S I G M A f i l ter . T a b l e 7.1: S u m m a r y of measured noise (VNCT) a n d ca lculated dose uncer ta inty (<7£)) a n d dose resolut ion w i t h 9 5 % confidence (D^%) for the unf i l tered a n d a l l f i l tered images. ^Performance of the S U S A N filter is independent of m a s k size. L o w Dose R e g i o n H i g h Dose R e g i o n F i l t e r M a s k C T J V c t (H) OD ( cGy) Df% ( cGy) &D ( cGy) Df% ( cGy) U n f i l t e r e d — 0.93 79 219 152 421 M E A N 3 x 3 0.38 33 90 63 173 5 x 5 0.27 23 63 44 121 7 x 7 0.20 17 46 32 89 M E D I A N 3 x 3 0.50 42 117 81 226 5 x 5 0.31 26 72 50 139 7 x 7 0.09 8 21 14 40 M I D P O I N T 3 x 3 0.49 42 117 81 224 5 x 5 0.48 41 113 78 216 7 x 7 0.39 33 92 64 178 A D A P T I V E 3 x 3 0.41 35 98 68 188 5 x 5 0.26 25 70 48 134 7 x 7 0.23 19 54 37 103 a - M E A N 3 x 3 0.41 35 97 68 188 5 x 5 0.30 25 69 49 136 7 x 7 0.24 20 55 39 108 S I G M A 3 x 3 0.63 53 147 103 285 5 x 5 0.56 48 133 92 255 7 x 7 0.54 46 127 88 244 S U S A N 3 x 3* 0.43 36 100 70 194 7.3.2 S p a t i a l D o s e I n t e g r i t y H i g h Dose R e g i o n Profi les Prof i les t h r o u g h the same l o c a t i o n of the H D R i n each f i l tered image are c o m p a r e d to the unf i l tered profile i n figures 7.13, 7.14 a n d 7.15 for the 3 x 3, 5 x 5 a n d 7 x 7 masks respectively. T h e S U S A N f i l ter result is shown i n figure 7.13. M a g n i f i e d views of the r ight h a n d H D R p e n u m b r a are also p r o v i d e d i n a l l three figures. I n general , increased s m o o t h i n g a n d d e g r a d a t i o n of the H D R peak a n d profi le shapes occurs w i t h increas ing mask size. T h e la t te r is most c lear ly evidenced by the m a g -ni f ied p e n u m b r a plots . T h e except i on is the S I G M A fi lter whose response appears near ly independent of m a s k size. T h e M E A N a n d a - M E A N filters prov ide v i r t u a l l y i d e n t i c a l results (especial ly for the 5 x 5 a n d 7 x 7 masks) a n d s m o o t h the profi le more t h a n the other f i lters. I n a d d i t i o n , a n a r r o w i n g of the H D R a n d spread ing of the p e n u m b r a is apparent for these f i l ters , especial ly for the 7 x 7 mask . T h e M E -D I A N a n d M I D P O I N T filters prov ide " s tep - l ike " s m o o t h i n g . T h e M E D I A N fi lter m a i n t a i n s sharp edges, p a r t i c u l a r l y i n the low dose wings , bet ter t h a n the M E A N f i l ter , however i t doesn 't appear to m o d e l the p e n u m b r a very we l l . T h e M I D P O I N T fi lter appears to per form e r r a t i c a l l y c o m p a r e d to the rest, c lear ly at t imes affected b y out l iers i n out of p lane pixels . It also d r a m a t i c a l l y deforms the peak shape for the 7 x 7 mask. T h e S U S A N , 3 x 3 m a s k A D A P T I V E a n d S I G M A filters per f o rm s i m i l a r l y a n d m a i n t a i n in tens i ty changes greater t h a n the average noise level better t h a n any of the other f i lters. T h i s is w e l l exempl i f i ed by the h i g h dose regions o n the peak. A s i l l u s t r a t e d i n the profi le p l o t s , for a l l m a s k sizes the A D A P T I V E fi lter best m a i n t a i n s the o r ig ina l peak shape, w i t h the S I G M A a n d S U S A N filters also per f o rming very wel l . However , at the larger masks the A D A P T I V E fi lter c lear ly s m o o t h the profi le more t h a n the S I G M A fi lter. 1500 <3 1000 w o Q 500 Unfiltered MEAN MEDIAN MIDPOINT ADAPTIVE alphaMEAN SIGMA SUSAN 20 40 60 80 Pixel Number 1000 9 0 0 £ 800 (/> o Q 700 600 500 100 120 98 MEAN Unfiltered .v. N \ > . \ MEDIAN \ \ \ \ MIDPOINT J ADAPTIVE I \ a-MEAN SIGMA SUSAN 99 100 101 102 103 Pixel Number F i g u r e 7.13: F u l l HDR profiles and a magn i f i ca t i on of the r ight h a n d p e n u m b r a (at r ight) for the unf i l tered image a n d filtered images for the 3 x 3 mask size. Pixel Number Pixel Number F i g u r e 7.14: F u l l H D R profiles a n d a magni f i ca t i on of the r ight h a n d p e n u m b r a (at r ight ) for the unf i l tered image a n d f i l tered images for the 5 x 5 mask size. Pixel number Pixel number F i g u r e 7.15: F u l l H D R profiles a n d a magn i f i ca t i on of the r ight h a n d p e n u m b r a (at r ight ) for the unf i l t ered image a n d f i l tered images for the 7 x 7 mask size. I n order to quant i fy the effects of the fi lters o n the dose prof i le , tab le 7.2 provides measures of peak w i d t h at the 5 0 % dose level (750 c G y ) a n d the w i d t h of the p e n u m b r a reg ion ( from 500 to 1000 c G y ) p l o t t e d i n figures 7.13 t h r o u g h 7.15. I n general , a l l the filters are able to preserve the w i d t h of the H D R to w i t h i n 0.5 m m . However , the p e n u m b r a w i d t h differentiates f i l ter per formance . P e n u m b r a w i d t h is w e l l preserved us ing the S U S A N fi lter a n d the other f i lters for the 3 x 3 mask , w i t h the M E A N , M E D I A N a n d a - M E A N filters r e s u l t i n g i n a s l ight reduc t i on . F o r most f i lters, increas ing mask size increases the r e s u l t i n g p e n u m b r a w i d t h . T h e except ions are the A D A P T I V E a n d S I G M A filters for w h i c h mask size has very l i t t l e effect on the p e n u m b r a w i d t h . T h i s effect is espec ia l ly d r a m a t i c for the M I D P O I N T fi lter w h i c h d is tor ts the p e n u m b r a w i d t h b y several m m us ing the 7 x 7 mask . T h e M E D I A N fi lter is u n u s u a l i n tha t , for a l l m a s k sizes, i t narrows the p e n u m b r a . O v e r a l l , the A D A P T I V E , S I G M A a n d S U S A N filters do the best j ob at preserv ing the s p a t i a l in tegr i ty of the dose profi le . D o s e A r e a H i s t o g r a m s D A H values: A i o o , Ago a n d A 5 0 are p r o v i d e d for each f i l tered image a n d the u n f i l -tered image i n table 7.3. I n general , f i l t e r ing reduces A^oo, a n effect t h a t increases for larger mask sizes. T h i s effect is most d r a m a t i c for the M E A N a n d a - M E A N f i l -ters w h i c h reduce Aioo to v i r t u a l l y zero even u s i n g a 3 x 3 mask. I n a d d i t i o n , w i t h the except ion of the M I D P O I N T f i l ter , a l l 5 x 5 a n d 7 x 7 mask f i lters also s ign i f i -cant ly reduce Aioo- A s such, these fi lters w o u l d not be appropr ia te i n cases where accurate measurement of s m a l l , h i g h l y l o c a l i z e d h i g h dose regions is requ ired . T h e 3 x 3 mask M E D I A N , A D A P T I V E a n d S I G M A filters as w e l l as the S U S A N fi lter p e r f o r m m u c h better , m a i n t a i n i n g at least some m a x i m u m dose area. These p a r t i a l reduct ions of A ioo m a y not necessari ly ind i ca te loss of c r i t i c a l h i g h dose i n f o r m a t i o n as noise w i l l c ontr ibute to the unf i l tered value . T h e M I D P O I N T f i l ter provides the largest A ioo a n d , i n c ompar i son to the other f i l ters, provides l i t t l e decrease i n A ioo T a b l e 7.2: S u m m a r y of H D R (at 750 c G y ) a n d r ight h a n d p e n u m b r a w i d t h (1000 to 500 c G y ) for the unf i l tered a n d a l l f i l tered images. ^Performance of the S U S A N f i l ter is independent of m a s k size. F i l t e r M a s k W i d t h ( m m ) P e n u m b r a ( m m ) U n f i l t e r e d — 32.8 2.8 M E A N 3 x 3 32.9 2.4 5 x 5 32.7 3.2 7 x 7 32.8 3.8 M E D I A N 3 x 3 33.0 2.2 5 x 5 32.4 2.5 7 x 7 32.4 2.5 M I D P O I N T 3 x 3 33.1 2.8 5 x 5 33.3 3.5 7 x 7 32.3 4.5 A D A P T I V E 3 x 3 32.9 2.8 5 x 5 32.8 2.9 7 x 7 32.8 2.9 a - M E A N 3 x 3 33.1 2.2 5 x 5 32.7 2.9 7 x 7 32.6 3.3 S I G M A 3 x 3 32.8 2.8 5 x 5 32.7 2.7 7 x 7 32.7 2.6 S U S A N 3 x 3t 32.9 2.7 Tab le 7.3: S u m m a r y of ca l cu la ted dose area h i s t o g r a m values for the unf i l tered a n d a l l f i l tered images. ^Performance of the S U S A N fi lter is independent of mask size. F i l t e r M a s k AW0(cm2) A 8 0 ( c m 2 ) A 5 0 ( c m ' i ) U n f i l t e r e d — 1.22 6.20 9.40 M E A N 3 x 3 0.07 6.39 9.53 5 x 5 0.00 6.09 9.67 7 x 7 0.00 5.71 9.75 M E D I A N 3 x 3 0.36 6.48 9.32 5 x 5 0.09 6.35 9.21 7 x 7 0.00 6.28 9.16 M I D P O I N T 3 x 3 0.58 6.15 9.77 5 x 5 0.55 5.63 10.07 7 x 7 0.46 4.95 10.36 A D A P T I V E 3 x 3 0.20 6.44 9.48 5 x 5 0.10 6.55 9.40 7 x 7 0.03 6.54 9.37 a - M E A N 3 x 3 0.07 6.50 9.57 5 x 5 0.00 6.24 9.52 7 x 7 0.00 6.09 9.57 S I G M A 3 x 3 0.23 6.41 9.40 5 x 5 0.11 6.52 9.40 7 x 7 0.11 6.50 9.39 S U S A N 3 x 3t 0.13 6.52 9.47 w i t h increas ing m a s k size. F o r the 3 x 3 m a s k size, a l l f i lters m a i n t a i n Aso a n d A50 w i t h i n 0.4 c m 2 of the unf i l tered value . O v e r a l l , the A D A P T I V E a n d S I G M A filters change A 8 o a n d A50 the least c o m p a r e d to the unf i l tered values. I n terms of the effect of filter size, the filters t h a t v i r t u a l l y e l iminate A i o o , ( M E A N a n d a - M E A N ) , decrease Aso a n d increase A50 w i t h increas ing m a s k size. T h i s is consistent w i t h the spread ing out of the h i g h dose reg ion as observed t h r o u g h the prof i le analys is . T h e M E D I A N filter consistent ly produces Aso larger t h a n , a n d A50 smal ler t h a n , the unf i l t ered image a n d decreases b o t h Aso a n d A50 w i t h increas ing m a s k size. T h e M I D P O I N T filter has the opposite overa l l effect: Aso is cons is tent ly smal ler t h a n , a n d A50 larger t h a n , the unf i l tered image. T h i s effect increases d r a m a t i c a l l y w i t h increas ing m a s k size. T h e A D A P T I V E a n d S I G M A filters show the least dependence o n m a s k size. 7.4 Chapter Summary T h i s chapter invest igated the p o t e n t i a l of d i g i t a l image filtering for i m p r o v i n g dose reso lut ion i n C T gel dos imetry . Noise i n C T images was f ound to be w e l l m o d e l l e d b y G a u s s i a n d i s t r i b u t i o n s a n d for the purposes of gel dos imetry , f ound to be inde -pendent of s ignal s t rength . A s such, s p a t i a l filters for r e m o v i n g add i t i ve , s p a t i a l l y invar iant noise were tested. Tes t ing was done o n a C T image of a gel i r r a d i a t e d w i t h a c l i n i c a l l y relevant dose d i s t r i b u t i o n . F i l t e r per formance was f ound to v a r y great ly w i t h respect to b o t h achieved dose resolut ions a n d affects o n the s p a t i a l d i s t r i b u t i o n of dose. F o r the dose d i s t r i b u t i o n e x a m i n e d here, the 3 x 3 m a s k size A D A P T I V E a n d S U S A N filters p r o v i d e d the best overa l l per formance , more t h a n h a l v i n g the dose reso lut ion w i t h o u t s ign i f i cant ly d i s t o r t i n g the s p a t i a l d i s t r i b u t i o n of dose. I n conc lus ion , f i l ter ing C T gel images has been s h o w n to prov ide a n appre -c iable improvement i n dose reso lut i on w i t h o u t s igni f i cant ly d i s t o r t i n g s p a t i a l dose i n f o r m a t i o n . A p o t e n t i a l a p p l i c a t i o n is for v o l u m e i m a g i n g where , as descr ibed i n chapter 6, use of h igher noise i m a g i n g protoco ls m a y be necessary i n order to o b t a i n results w i t h i n a reasonable i m a g i n g t i m e . A s a n a d d i t i o n a l note, th i s work was one of the first to focus o n image f i l t e r ing as a t o o l to improve d o s i m e t r y a n d other techniques w i t h noise l i m i t a t i o n s , such as M o n t e C a r l o dos imetry , m a y benefit f r om th i s work . Chapter 8 Conclusions T h i s thesis has invest igated the use of x - r a y C T for e x t r a c t i n g dose i n f o r m a t i o n f r o m i r r a d i a t e d p o l y a c r y l a m i d e gels. T h e d r i v i n g force for th i s w o r k was the promise t h a t C T gel dos imetry has shown as a p r a c t i c a l 3 D t o o l for r a d i a t i o n therapy dose ver i f i cat ion a n d the m a i n barr ier to i t s i m p l e m e n t a t i o n , i t s low dose reso lut ion . T h e w o r k was d i v i d e d into three studies: a s t u d y of gel c o m p o s i t i o n , a s t u d y of sys tem noise a n d dose reso lut i on a n d a d i g i t a l image f i l t e r ing study. T h e results f r om a l l three studies are s u m m a r i z e d i n sect ion 8.1 a n d sect ion 8.2 discusses p o t e n t i a l d irect ions for future work . 8.1 Summary of Results T h e first study, the c o m p o s i t i o n a l s tudy, consisted of three components . T h e first component invo lved i n i t i a l studies to des ign a n d test a new sys tem for C T i m a g i n g gel v ia ls a n d to measure dose response reproduc ib i l i t y . T h e designed v i a l i m a g -i n g system used a s tyro foam p h a n t o m a n d was f ound to prov ide a n ~ 11 t imes improvement i n S N R over previous designs w h i l e s t i l l m a i n t a i n i n g u n i f o r m image in tens i ty w i t h i n the v ia l s . T h e i n t r a - b a t c h a n d i n t e r - b a t c h r e p r o d u c i b i l i t y of the gel C T dose response were b o t h f ound to be excel lent , v a l i d a t i n g the gel manufac ture , i r r a d i a t i o n , i m a g i n g a n d d a t a analys i s techniques used i n th is study. T h e excel lent i n t e r - b a t c h r e p r o d u c i b i l i t y also h i g h l i g h t e d the p o t e n t i a l for es tab l i sh ing s t a n d a r d -i zed c a l i b r a t i o n curves i n C T gel dos imetry . T h e second component of th is s t u d y character ized C T dose response for different gel compos i t ions , i n p a r t i c u l a r for gels w i t h v a r y i n g % C . R e s u l t s showed a large v a r i a t i o n i n b o t h the f u n c t i o n a l f o r m a n d the s t rength of the C T dose response w i t h % C . T h e observed t r e n d t h a t m i d - r a n g e % C gels show a larger a n d less l inear response t h a n the h i g h a n d low % C gels was s i m i l a r to tha t shown i n the l i t e ra ture t h r o u g h M R I a n d R a m a n spectroscopic a n a l -ysis. Dose response sens i t iv i ty a n d dose range were ana lysed a n d results ind i ca te t h a t b o t h factors are affected by gel % C a n d must be considered w h e n select ing a n o p t i m u m gel f o r m u l a t i o n for a p a r t i c u l a r a p p l i c a t i o n . I n p a r t i c u l a r , a 100 % C gel was shown to prov ide a n extended dose range, out to 100 G y , t h a t m a y f ind use i n c l i n i c a l app l i ca t i ons such as h i g h dose rate brachytherapy . T h e f ina l component of th i s s t u d y invest igated the effect of gel c o m p o s i t i o n on the r a d i a t i o n i n d u c e d dens i ty change t h a t occurs i n the gels a n d provides the contrast r equ i red for C T read-out . A m o d e l was developed to describe gel dens i ty change as a f u n c t i o n of p o l y m e r y i e l d a n d a n in t r ins i c dens i ty change that occurs o n conversion of m o n o m e r to p o l y m e r . B y a p p l y i n g th is m o d e l to measured C T dose response d a t a a n d i n c o r p o r a t i n g R a -m a n spectroscopic d a t a f rom the l i t e r a t u r e , two interest ing f u n d a m e n t a l propert ies of the gel densi ty to dose response were discovered. T h e first p r o p e r t y is t h a t the i n t r i n s i c densi ty change depends on gel % C . I n p a r t i c u l a r i t was discovered t h a t low a n d h i g h % C gels show a n ~ 1.5 t imes greater i n t r i n s i c dens i ty change t h a n intermediate % C gels. T h i s t r e n d is oppos i te to p o l y m e r y i e l d w h i c h is greatest for in termediate % C gels. It is h y p o t h e s i z e d t h a t the increased dens i ty i n the low a n d h i g h % C gels contr ibutes to the reduced y i e l d . T h e second p r o p e r t y is tha t for gels w i t h a g iven % C , the dens i ty change is l inear w i t h gel % T . F r o m the mode l , the inference is that b o t h the in t r ins i c dens i ty change a n d the f rac t i on of m o n o m e r c onsumed w i t h dose are solely dependent o n gel %C. F r o m a c l i n i c a l s ta ndpo in t , th i s result means that increas ing % T is a p o t e n t i a l means of increas ing gel sens i t iv i ty to C T read-out . I n s u m m a r y , these c o m p o s i t i o n a l studies (1) es tab l i shed a n i m p r o v e d m e t h o d of i m a g i n g gel v ia ls a n d the r e p r o d u c i b i l i t y of gel C T dose response, (2) defined the effect of gel % C o n dosimeter sens i t iv i ty a n d dose range a n d (3) deve l -oped a m o d e l to describe gel dens i ty change t h a t was successfully used to discover f u n d a m e n t a l propert ies of the response of gel dens i ty to dose. T h e second set of studies d e t e r m i n e d strategies for m i n i m i z i n g noise i n C T gel dos imetry a n d assessed system per formance . P h a n t o m design, C T i m a g i n g tech-n ique a n d voxe l d imensions were each invest igated a n d a l l were demonst ra ted to be i m p o r t a n t factors i n noise r educ t i on . Noise increases e x p o n e n t i a l l y w i t h p h a n t o m d iameter a n d the design of s m a l l p h a n t o m s is r e c o m m e n d e d wherever possible . I n terms of selectable scan parameters , choice of r e cons t ruc t i on a l g o r i t h m has the most d r a m a t i c effect o n image noise a n d a lgor i thms designed to h igh l ight h i g h frequency i n f o r m a t i o n (edges, fine tissue i n l u n g etc.) s h o u l d be avo ided for gel dos imetry . Increasing x - r a y tube voltage, current a n d slice scan t i m e a l l reduce image noise, however increas ing t u b e voltage was shown to be most efficient a n d s h o u l d be p r i -o r i t i z e d w h e n select ing a s canning pro toco l . F o l l o w i n g th i s , t u b e current a n d slice scan t i m e can be increased, however b o t h involve a compromise w i t h i m a g i n g t ime . Image noise decreases w i t h increas ing voxel d imensions , a n effect t h a t was d e m o n -s t ra ted to be more d r a m a t i c for p i x e l size t h a n for slice th ickness . A s a result , larger voxe l sizes (for example 2 x 2 x 3 m m 3 ) s h o u l d be considered i n some s i tuat ions . I n a d d i t i o n , these studies i l l u m i n a t e a n a b i l i t y t o pred i c t image noise for any g iven scan pro to co l g iven noise measured for a single reference pro toco l . T h e compromise between i m a g i n g t i m e a n d image noise was invest igated i n d e t a i l a n d quant i f i ed for vo lume i m a g i n g over a range of voxe l d imensions . G i v e n a k n o w n gel dose response, these i m a g i n g t i m e results a long w i t h the a b i l i t y to pred i c t image noise al lows for the p r e d i c t i o n of b o t h i m a g i n g t ime a n d dose reso lut i on for a g iven scan pro toco l . T h i s is a va luable p r a c t i c a l result since i t means t h a t a n i m a g i n g p r o t o c o l for C T gel dos imetry c a n now be selected based o n c l i n i c a l requirements . F i n a l l y , the results of these studies were c o m b i n e d to prov ide measures of achievable dose reso lut i on for the P A G formulat ions s tud ied i n chapter 5. T h i s was done over a range of voxe l sizes for b o t h single slice a n d vo lume i m a g i n g , g iven a 1 hour t i m e const ra int . F o r single slice i m a g i n g , dose resolut ions (-DA,%) of 5 % a n d 2.2 % can be achieved us ing voxe l sizes of 1 x 1 x 3 m m 3 a n d 2.5 x 2.5 x 3.0 m m 3 , respectively. F o r vo lume i m a g i n g , g iven a 1 hour t i m e l i m i t , the result is not as good , but i t is encourag ing t h a t a dose reso lut i on of < 5% c a n be achieved u s i n g the 2.5 x 2.5 x 3.0 m m 3 voxe l size. However , u s i n g a smal ler voxe l size is unacceptab le for vo lume i m a g i n g : % is 12 % given a 1 x 1 x 3 m m 3 voxe l size. I n s u m m a r y , these studies d e t e r m i n e d o p t i m i z a t i o n strategies for m i n i m i z i n g noise i n C T gel dos imetry a n d quant i f i ed sys tem per formance . A compar i son w i t h other gel dos imetry results f r o m recent l i t e ra ture indicates t h a t dose reso lut i on is s t i l l not at the level achievable w i t h M R I or O C T read-out . However , th i s w o r k has made s igni f icant improvement i n the dose reso lut ion achievable w i t h C T gel d o s i m e t r y a n d , w i t h i m a g i n g t i m e at less t h a n 1 m i n u t e for a single slice, the technique is faster t h a n b o t h current M R I a n d O C T techniques. T h e t h i r d s t u d y invest igated the p o t e n t i a l of d i g i t a l f i l t e r ing for r e d u c i n g noise i n gel C T images a n d hence i m p r o v i n g dose reso lut ion . T h e s t u d y h a d two components : a charac ter i za t i on of C T image noise a n d a compar i son of f i l ter perfor -mance . C T image noise was character ized by m e a s u r i n g p r o b a b i l i t y dens i ty func -t ions for b o t h a dens i ty p h a n t o m a n d gel images. It was f ound to be G a u s s i a n d i s t r i b u t e d a n d for the purposes of gel dos imetry , independent of image contrast level (or s igna l s t rength) . A s such, s p a t i a l f i lters for r emov ing add i t i ve , s p a t i a l l y invar iant noise were used. F i l t e r s were tested i n the i r a b i l i t y to improve dose res-o l u t i o n a n d i n the i r affect on the s p a t i a l d i s t r i b u t i o n of dose. F i l t e r per formance was f ound to v a r y greatly. T h e M E A N fi lter was best at i m p r o v i n g dose reso lu t i on b u t th is improvement was at the expense of m a i n t a i n i n g the s p a t i a l d i s t r i b u t i o n of dose. T h e a - M E A N a n d A D A P T I V E f i lters p r o v i d e d dose reso lut ions at a l l m a s k sizes tha t was comparab le to , b u t s l i ght ly worse t h a n , the M E A N f i l ter . I n a d d i t i o n , the S U S A N filter m a t c h e d these i n per formance for the 3 x 3 m a s k sizes. I n terms of the a b i l i t y of the filters to m a i n t a i n the s p a t i a l in tegr i ty of the unf i l t ered dose d i s t r i b u t i o n , the best overal l results were ob ta ined u s i n g the A D A P T I V E , S I G M A a n d S U S A N filters. T h e M E D I A N f i l ter was also re la t ive ly good at m a i n t a i n i n g s p a t i a l dose in tegr i ty except at ex t remely h i g h dose gradients . F o r the c l i n i c a l dose d i s t r i b u t i o n tested, the A D A P T I V E a n d S U S A N filters p r o v i d e d the best overa l l per formance . U s i n g just a s m a l l , 3 x 3 p i x e l mask A D A P T I V E filter or the S U -S A N filter, dose reso lut ion was m o r e t h a n ha lved w i t h o u t s ign i f i cant ly d i s t o r t i n g the s p a t i a l d i s t r i b u t i o n of dose. I n s u m m a r y , th is s t u d y showed t h a t noise i n gel C T images is G a u s s i a n d i s t r i b u t e d a n d s p a t i a l l y invar ient a n d d e m o n s t r a t e d t h a t image filtering is a n effective t o o l for i m p r o v i n g dose reso lut ion i n C T P A G dos imetry . I n s u m m a r y , th is thesis has s ign i f i cant ly fur thered the field of gel dos ime-t r y a n d has paved the way for a very pos i t ive c l i n i c a l future for the use of x - r a y C T as a read-out t oo l for gel dos imetry . N e w ins ight has been ga ined into the f u n d a m e n t a l nature of P A G dens i ty to dose response w h i c h w i l l a i d i n future gel development efforts. O p t i m i z a t i o n strategies have been developed for m i n i m i z i n g noise i n C T gel dos imetry a n d achievable dose reso lut i on g iven i m a g i n g t i m e a n d voxe l size constra ints has been i m p r o v e d a n d quant i f ied . F i n a l l y , d i g i t a l image fil-t e r i n g has been demonstrated to be a n effective t o o l for fur ther improvements to dose reso lut ion . W h e n these advances are considered together w i t h the access ib i l i ty a n d low cost of C T i m a g i n g the p o t e n t i a l for C T gel dos imetry to become a c l in i ca l successful 3 D dosimeter is clear. However , cont inued w o r k is needed, i n p a r t i c u l a r i n gel development , a n d possible d irect ions are descr ibed below. 8.2 Future Work T h e r e are several p o t e n t i a l avenues for extens ion of th i s work . T h e first is i n the area of gel f o r m u l a t i o n development a n d o p t i m i z a t i o n . T h i s w o r k has i l l u m i n a t e d re lat ionships between gel f o r m u l a t i o n a n d C T dose response w h i c h w i l l be useful i n searching for a d d i t i o n a l p o l y m e r gel systems that m a y prov ide enhanced response to C T read-out . T h i s avenue w i l l also focus i n a new d i r e c t i o n , o n n o r m o x i c gels w h i c h can be made o n a bench- top , i n the presence of oxygen. I n a d d i t i o n , a c ry -l a m i d e is tox ic a n d o p t i m i z a t i o n of future gel f o rmulat i ons for C T gel dos imetry w i l l concentrate on non - tox i c monomers . Secondly, th i s w o r k has focused o n s i n -gle slice C T scanning since these are the types of scanners c u r r e n t l y avai lable for C T s i m u l a t i o n i n most cancer hospi ta ls . However , m u l t i - s l i c e C T scanners are a l -ready prevalent a n d w i l l surely be c o m m o n for C T s i m u l a t i o n i n the near future . F u t u r e w o r k w i l l invest igate the use of m u l t i - s l i c e C T systems for gel dos imetry . A n o t h e r interest ing p o t e n t i a l for future w o r k is the ex tens ion of the f i l ter ing w o r k into another field, M o n t e C a r l o dos imetry . M o n t e C a r l o tools are cur rent ly under development to m o d e l dose d i s t r ibut i ons de l ivered by l inacs i n pat ient a n a t o m y ( C T images) . However , even w i t h l ong c a l c u l a t i o n t imes , the dose d i s t r i b u t i o n s p r o d u c e d are quite no isy a n d there is interest i n e x p l o r i n g f i l t e r ing techniques for i m p r o v e d dose reso lut ion , s i m i l a r to the w o r k presented i n th is thesis . Appendix A Digital Image Filters T h e fo l lowing is a descr ip t ion of each of the d i g i t a l image filters invest igated i n th is work . T h e filters are a l l based o n the same general p r i n c i p l e : a mask ( m x n pixels ) is centered on each p i x e l of the image , g(s, t), a n d a f u n c t i o n is app l i ed to the image pixels i n the region of the mask (Sxy) so t h a t the center p i x e l is replaced w i t h a new value, f(x,y). M e a n F i l t e r T h e mean , or "averaging" or " box " , filter ( M E A N ) is the s implest a n d most c o m m o n noise r e d u c t i o n filter [209]. It replaces each image p i x e l w i t h the average value of p ixels i n Sxy. T h e M E A N filter is g iven by : (sjt) € Sxy A m e a n filter provides good image s m o o t h i n g b u t is k n o w n to b l u r edges. It is i n c l u d e d i n th i s invest igat ion large ly as a po in t of reference. M e d i a n F i l t e r T h e m e d i a n filter ( M E D I A N ) is the most we l l k n o w n filter based on order -s tat i s t i cs . It operates i n a s imi lar fashion to a m e a n filter except t h a t each p i x e l is rep laced (A .1 ) b y the m e d i a n , rather t h a n the m e a n , of the pixels i n Sxy. T h e M E D I A N filter is g iven by : f(x,y) = median[Stt)eSxy{g{s,t)} (A .2) M e d i a n fi lters are popu lar for the i r a b i l i t y to preserve edges better t h a n m e a n fi lters wh i l e s t i l l effectively reduc ing noise. M i d p o i n t F i l t e r T h e m i d p o i n t or midrange f i lter ( M I D P O I N T ) is another order -s tat i s t i cs f i l ter . It replaces each image p i x e l w i t h the m i d p o i n t value between the m a x i m u m a n d m i n -i m u m values i n Sxy. T h e M I D P O I N T fi lter is g iven by: f(x,y) = ^[rnax{Stt)esxy{g{s,t)} + min{Stt)eSxy{g(s,t)}} (A .3) M i d p o i n t niters per f o rm best for r a n d o m l y d i s t r i b u t e d noise, such as G a u s s i a n noise [206]. M e a n A d a p t i v e F i l t e r A d a p t i v e filters differ f r om the prev ious fi lters descr ibed here i n t h a t the i r behav iour changes based on the l o ca l s t a t i s t i c a l character is t i cs measured i n Sxy. T h e response of a m e a n adapt ive fi lter ( A D A P T I V E ) is based o n a c o m p a r i s o n of these l o c a l s tat ist ics w i t h the overa l l image noise. T h e A D A P T I V E fi lter is g iven by : f(x,y)=g(s,t)-^[g(s,t)-mL] (A .4) where mi, a n d a\ are the l o c a l m e a n a n d var iance i n Sxy a n d a\ is the var iance of the noise c o r r u p t i n g the image. T y p i c a l l y , a\ is t a k e n as the var iance measured for a region of the image w i t h no contrast . T h e filter adapts i t s response to the value of o\ c ompared to a\ such t h a t i f cr£ > > cr^, a value close t o the o r i g i n a l value , g(s, t), is re turned . T h i s provides a means of preserv ing edges [206]. A l p h a - T r i m m e d M e a n F i l t e r T h e a l p h a - t r i m m e d m e a n fi lter ( a - M E A N ) combines order -s tat i s t i cs a n d a m e a n filter. It replaces each p i x e l w i t h the m e a n value of i t s ne ighbours i n Sxy trimmed of a g iven f ract ion , a l p h a (a ) , of i ts lowest a n d highest values. F o r example , for a = 0.25, the quarter lowest a n d highest values i n SXy w i l l be exc luded i n c a l c u l a t i o n of the mean . O n e can see t h a t the a l p h a - t r i m m e d m e a n filter spans a c o n t i n u u m of filters whose e x t r e m a are given b y the m e a n filter (a = 0) a n d the m e d i a n filter ( a = 0.5) [210]. T h r o u g h o u t th is work a = 0.25 was used. T h e a - M E A N f i l ter is g iven by: f(x,v) = A TT E 9r(s,t) (A .5) mn — 2 a m n - 1) V ; (s,t)eSxy A l p h a - t r i m m e d m e a n filters are p a r t i c u l a r l y useful i n cases where a n image is de-graded b y m u l t i p l e types of noise [206]. S i g m a M e a n F i l t e r T h e s i g m a m e a n filter ( S I G M A ) is s i m i l a r to a n a - M E A N filter i n i t l i m i t s the pixels used to ca lculate the m e a n to those "close t o " the centre p i x e l value. It uses the s i g m a (a) f r om G a u s s i a n d i s t r i b u t e d noise a n d calculates the m e a n u s i n g o n l y those pixels i n the mask t h a t are w i t h i n a fixed s i g m a range f r o m the centre p i x e l . T y p i c a l l y , a range of 2a is used [211]. I n th is way, the filter on ly averages over those pixels of s i m i l a r in tens i ty to the centre p i x e l a n d provides a means of preserv ing edges a n d de ta i l i n the image. T h e S I G M A filter is g iven by: f { x ' y ) = Nrhs& £ 9 { s ' t ] ( A - 6 ) where S2y is those mask elements w i t h i n 2a of t h e centre p i x e l intensity . T h e S I G M A fi lter as descr ibed above does not remove spot noise. T h i s p r o b l e m is e l i m i n a t e d b y re f in ing the filter so t h a t i f e Sxy < K t h e n the m e a n of the eight nearest neighbours is c a l c u l a t e d instead of f(x,y) as above. T h e recommended value of i f is 2 a n d 3 for 5 x 5 a n d 7 x 7 masks respect ive ly [211]. Smallest U n i v a l u e Segment A s s i m i l a t i n g N u c l e u s F i l t e r Smal lest U n i v a l u e Segment A s s i m i l a t i n g Nuc leus ( S U S A N ) is a new a p p r o a c h to image processing, i n p a r t i c u l a r , to edge a n d corner detec t ion a n d noise r e d u c t i o n [207]. T h e basis of the S U S A N p r i n c i p l e is s i m i l a r to the S I G M A fi lter i n t h a t each image p i x e l (the Nuc leus ) has assoc iated w i t h i t a l o c a l area of s i m i l a r in tens i ty (the U S A N ) . Differences inc lude G a u s s i a n based s m o o t h i n g i n the br ightness d o m a i n a n d r e m o v i n g the centre p i x e l i tsel f f r om ca l cu lat ions . T h i s f i lter c la ims to inc lude the best propert ies of, a n d to ou tper f o rm, m a n y edge detec t ion s m o o t h i n g a lgor i thms . T h e S U S A N noise r e d u c t i o n a l g o r i t h m is g iven by: _ ^ _ ( » ( « . t ) - / ( x , » ) ) a f(x'y}= ± - T2 ( f l ( . , t ) - / ( ^ p ( A - 7 ) where S~y is Sxy e x c l u d i n g the center p i x e l , r is the p i x e l rad ius f r o m the centre p i x e l a n d t is the br ightness thresho ld parameter . F i l t e r per formance is re la t ive ly independent of t ( this parameter is ra re ly varied) a n d is independent of mask size [207]. T h r o u g h o u t th is w o r k t was set to a r e commended value of 2a [207]. Bibliography [1] N C I C . C a n a d i a n cancer s tat is t i cs . T e c h n i c a l report , N a t i o n a l C a n c e r Ins t i tu te of C a n a d a , 2004. [2] C . A . Perez a n d L . W . B r a d y , edi tors . Principles and Practice of Radiation Oncology. L i p p i n c o t t - R a v e n , P h i l i d e l p h i a , 1998. [3] A . J . N e a l a n d P . J . H o s k i n . Clinical Oncology: Basic Principles and Practice. O x f o r d U n i v e r s i t y Press , N e w Y o r k , 2 n d e d i t i o n , 1997. [4] D . W . M i l l e r . A review of p r o t o n b e a m r a d i a t i o n therapy. Med. Phys., 22(11): 1943-54, 1995. [5] J . Debus , O . J a c k e l , G . K r a f t , a n d M . W a n n e n m a c h e r . Is there a role for heavy i o n b e a m therapy? Recent Results Cancer Res, 150:170-82, 1998. [6] F . M . K h a n . The Physics of Radiation Therapy. W i l l i a m s a n d W i l k i n s , B a l -t imore , M D , 2 n d e d i t i o n , 1994. [7] J . V a n D y k , editor . The Modern Technology of Radiation Oncology: A Com-pendium for Medical Physicists and Radiation Oncologists. M e d i c a l P h y s i c s P u b l i s h i n g , M a d i s o n , W I , 2000. [8] S. W e b b . The Physics of Three-Dimensional Radiation Therapy: Conformal Radiotherapy, Radiosurgery, and Treatment Planning. I n s i t i t u t e of P h y s i c s P u b l i s h i n g , B r i s t o l , U K , 1993. [9] B . H . D i d i n g e r , W . Schlegel , a n d J . D e b u s . I n t e n s i t y - m o d u l a t e d rad i o therapy - technology a n d c l i n i c a l app l i ca t i ons . Onkologie, 25(3) :233-8, 2002. [10] G . A . E z z e l l , J . M . G a l v i n , D . L o w , J . R . P a l t a , I. R o s e n , M . B . Sharpe , P . X i a , Y . X i a o , L . X i n g , a n d C . X . Y u . G u i d a n c e document o n del ivery, t reatment p l a n n i n g , a n d c l i n i c a l i m p l e m e n t a t i o n of I M R T : report of the I M R T S u b c o m m i t t e e of the A A P M R a d i a t i o n T h e r a p y C o m m i t t e e . Med. Phys., 30 (8) :2089-115, 2003. [11] S. P o r c e d d u , G . H o p e , J . W i l l s , J . C r a m b , D . R i s c h i n , J . C o r r y , a n d L . Peters . I n t e n s i t y - m o d u l a t e d rad io therapy : examples of i t s u t i l i t y i n head a n d neck cancer. Australas. Radiol., 4 8 ( l ) : 5 1 - 7 , 2004. [12] H . E . J o h n s a n d J . R . C u n n i n g h a m . The Physics of Radiology. C h a r l e s C T h o m a s P u b l i s h e r , Spr ing f ie ld , I L , 1983. [13] F . H . A t t i x . Introduction to Radiological Physics and Radiation Dosimetry. J o h n W i l e y a n d Sons, N e w Y o r k , N Y , 1986. [14] M . O l d h a m , J . H . Siewerdsen, A . Shetty , a n d D . A . Jaffray. H i g h reso lut i on gel dos imetry by o p t i c a l - C T a n d M R scann ing . Med. Phys., 28(7) :1436-45, 2001. [15] I C R U . D e t e r m i n a t i o n of absorbed dose i n a pat ient i r r a d i a t e d by beams of x or g a m m a rays i n rad i o therapy procedures . T e c h n i c a l report , I n t e r n a t i o n a l C o m m i s s i o n on R a d i a t i o n U n i t s M e a s u r e m e n t ( I C R U ) , 1979. [16] P . R . A l m o n d , P . J . B iggs , B . M . Coursey , W . F . H a n s o n , M . S. H u q , R . N a t h , a n d D . W . O . Rogers . A A P M ' s T G - 5 1 p r o t o c o l for c l i n i c a l reference d o s i m e t r y of h igh-energy p h o t o n a n d e lectron beams. Med. Phys., 2 6 ( l l ) : 1 8 4 7 - 7 0 , 1999. [17] A m e r i c a n A s s o c i a t i o n of P h y s i c i s t s i n M e d i c i n e R T C T a s k G r o u p 21. A p r o -t o co l for the d e t e r m i n a t i o n of absorbed dose f r o m high-energy p h o t o n a n d e lectron beams. Med. Phys., 10(6) :741-71, 1983. [18] F . M . K h a n , K . P . D o p p k e , K . R . H o g s t r o m , G . J . K u t c h e r , R . N a t h , S. C . P r a s a d , J . A . P u r d y , M . Rozen fe ld , a n d B . L . Werner . C l i n i c a l e l e c t ron -beam dos imetry : R e p o r t of A A P M R a d i a t i o n T h e r a p y C o m m i t t e e T a s k G r o u p N o . 25. Med. Phys., 18(1) :73-109, 1990. [19] W . U . L a u b a n d T . W o n g . T h e v o l u m e effect of detectors i n the d o s i m e t r y of s m a l l fields used i n I M R T . Med. Phys., 30(3) :341-7, 2003. [20] R . R a m a n i , S. R u s s e l l , a n d P . O ' B r i e n . C l i n i c a l d o s i m e t r y u s i n g M O S F E T s . Int. J. Rad. Oncol. Biol. Phys., 37(4) :959-64, 1997. [21] X . R . Z h u , J . J . A l l e n , J . S h i , a n d W . E . S i m o n . T o t a l scatter factors a n d t issue m a x i m u m rat ios for s m a l l rad iosurgery fields: c ompar i son of d iode detectors, a para l l e l -p la te i o n chamber , a n d r a d i o g r a p h i c f i l m . Med. Phys., 27(3) :472-7, 2000. [22] C . M a r t e n s , I. C laeys , C . D e W a g t e r , a n d W . D e Neve . T h e value of r a -d iographic f i l m for the charac te r i za t i on of i n t e n s i t y - m o d u l a t e d beams. Phys. Med. Biol., 47(13) :2221-34, 2002. [23] I. J . Yeo , A . B e i k i - A r d a k a n i , Y . B . C h o , M . H e y d a r i a n , T . Z h a n g , a n d M . Is-l a m . E D R 2 f i l m d o s i m e t r y for I M R T ver i f i cat ion u s i n g low-energy p h o t o n fi lters. Med. Phys., 31(7) :1960-3, 2004. [24] T . A c k e r l y , S. T o d d , I. W i l l i a m s , M . Geso , a n d J . C r a m b . C o m p a r a t i v e eva lua -t i o n of Wel lhofer i o n chamber ar ray a n d S c a n d i t r o n i x d iode a r r a y for d y n a m i c wedge dosimetry . Australas. Phys. Eng. Sci. Med., 20(2) :71-83, 1997. [25] A . E r t l , R . F . H a r t l , M . Zehetmayer , K . K i t z , a n d W . G r i f f i t t . T L D a r r a y for precise dose measurements i n stereotact ic r a d i a t i o n techniques. Phys. Med. Biol, 41(12) :2679-86, 1996. [26] D . L e t o u r n e a u , M . G u l a m , D . Y a n , M . O l d h a m , a n d J . W . W o n g . E v a l u a t i o n of a 2 D diode ar ray for I M R T q u a l i t y assurance. Radiother. Oncol., 70(2): 199-206, 2004. [27] J . L . R o b a r . Film Dosimetry and Verification of Three-Dimensional Con-formal Dose Distributions in Stereotactic Radiosurgery. P h D , U n i v e r s i t y of B r i t i s h C o l u m b i a , 2000. [28] J . L . R o b a r a n d B . G . C l a r k . A p r a c t i c a l technique for ver i f i cat ion of three d imens iona l dose d i s t r i b u t i o n s i n stereotact ic radiotherapy . Med. Phys., 27 (5) :978-87, 2000. [29] H . Fr i cke a n d S M o r s e . T h e chemica l a c t i o n of roentgen rays o n d i l u t e ferrous sulfate so lut ions as a measure of r a d i a t i o n dose. Am. J. Roentgenol. Radium Therapy Nucl. Med., 18:430-32, 1927. [30] J . C . Gore , Y . S. K a n g , a n d R . J . Schulz . M e a s u r e m e n t of r a d i a t i o n dose d i s t r ibut i ons b y nuclear magnet i c resonance ( N M R ) i m a g i n g . Phys. Med. Biol, 29:1189-97, 1984. [31] M . J . M a r y a n s k i , J . C . G o r e , R . P . K e n n a n , a n d R . Schulz . N M R r e l a x a t i o n enhancement i n gels p o l y m e r i z e d a n d cross - l inked b y i o n i z i n g r a d i a t i o n : a new approach to 3 -D dos imetry by M R I . Magn. Reson. Imaging, 11:253-258, 1993. [32] Y . D e Deene, C . D e W a g t e r , W . D e Neve , a n d E . A c h t e n . A r t e f a c t s i n m u l t i -echo T 2 i m a g i n g for h igh -prec i s i on gel dos imetry : I. analys is a n d compensat i on of eddy currents . Phys. Med. Biol, 45(7): 1807-23 , 2000. [33] Y . D e Deene, C . D e W a g t e r , W . D e Neve , a n d E . A c h t e n . A r t e f a c t s i n m u l t i -echo T 2 i m a g i n g for h igh -prec i s i on gel dos imetry : II . analys is of B l - f i e l d i n -homogeneity. Phys. Med. Biol, 45(7) :1825-39, 2000. [34] Y . D e Deene a n d C . D e Wagter . A r t e f a c t s i n mul t i - e cho T 2 i m a g i n g for h i g h -prec is ion gel dos imetry : I I I . effects of t empera ture dri f t d u r i n g s canning . Phys. Med. Biol., 46(10) :2697-711, 2001. [35] M . J . M a r y a n s k i , C . A u d e t , a n d J . C . G o r e . Effects of c ross l ink ing a n d t e m -perature o n the dose response of a B A N G p o l y m e r gel dosimeter . Phys. Med: Biol, 42 :303-311 , 1997. [36] J . C . G o r e , M . R a n d e , M . J . M a r y a n s k i , a n d R . J . Schulz . R a d i a t i o n dose d i s t r i b u t i o n s i n three d imensions f r o m tomograph i c o p t i c a l dens i ty s c a n n i n g of p o l y m e r gels: I Deve lopment of a n o p t i c a l scanner. Phys. Med. Biol., 41: 2695-704, 1996. [37] M . J . M a r y a n s k i , Y . Z. Zastavker , a n d J . C . G o r e . R a d i a t i o n dose d i s t r i b u t i o n s i n three d imensions f r om t o m o g r a p h i c o p t i c a l dens i ty s canning of p o l y m e r gels: I I . O p t i c a l propert ies of the B A N G p o l y m e r gel . Phys. Med. Biol., 41 (12) :2705-17, 1996. [38] C . B a l d o c k , L . R i n t o u l , S. F . K e e v i l , J . M . P o p e , a n d G . A . George . F o u r i e r t r a n s f o r m R a m a n spectroscopy of p o l y a c r y l a m i d e gels ( P A G s ) for r a d i a t i o n dos imetry . Phys. Med. Biol., 43 :3617-27, 1998. [39] M . H i l t s , C . A u d e t , C . D u z e n l i , a n d A . J i rasek . P o l y m e r gel d o s i m e t r y u s i n g x - r a y c o m p u t e d t omography : a f eas ib i l i ty study. Phys. Med. Biol., 45(9): 2559-71 , 2000. [40] M . L . M a t h e r , A . K . W h i t t a k e r , a n d C . B a l d o c k . U l t r a s o u n d e v a l u a t i o n of p o l y m e r gel dosimeters . Phys. Med. Biol, 47 :1449-58, 2002. [41] C . A u d e t , M . H i l t s , A . J i rasek , a n d C . D u z e n l i . C T gel d o s i m e t r y technique : c o m p a r i s o n of a p l a n n e d a n d measured 3 D stereotact ic dose vo lume . J. App. Clin. Med. Phys., 3 :110-118, 2002. [42] M . H i l t s . X-ray Computed Tomography for Performing Polymer Gel Dosime-try: A Feasibility Study. M S c , U n i v e r s i t y of B r i t i s h C o l u m b i a , 1999. [43] J . V . T r a p p , S. A . B a c k , M . Lepage , G . M i c h a e l , a n d C . B a l d o c k . A n ex-p e r i m e n t a l s t u d y of the dose response of p o l y m e r gel dosimeters i m a g e d w i t h x - r a y c o m p u t e d tomography. Phys. Med. Biol., 46(11) :2939-51, 2001. [44] W . R . Hendee a n d E . R . R i t e n o u r . Medical Imaging Physics. W i l e y - L i s s , N e w Y o r k , N Y , 4 t h ed i t i on , 2002. [45] T . S. C u r r y , J . E . Dowdey , a n d R . C . M u r r y . Christensen's Physics of Diag-nostic Radiology. L e a a n d Febiger , P h i l a d e l p h i a , P A , 1990. [46] B e u t e l J . , K u n d e l H . L . , a n d V a n M e t t e r R . L . Handbook of Medical Imaging, Volume 1: Physics and Psychophysics. S P I E Press , 2000. [47] J . H s i e h . Computed Tomography: Principles, Design, Artifacts and Recent Advances. S P I E Press , B e l l i n g h a m W A , 2003. [48] A . M . C o r m a c k . R e p r e s e n t a t i o n of a f u n c t i o n b y i ts l ine integrals , w i t h some rad io log i ca l app l i ca t i ons . Journal of Applied Physics, 34 :2722-2727, 1963. [49] G . N . Hounsf ie ld . C o m p u t e r i s e d transverse a x i a l s cann ing ( t omography) . P a r t I: D e s c r i p t i o n of system. Br. J. of Radiol., 46 :1016-22 , 1973. [50] C . H . M c C o l l o u g h a n d R . L . M o r i n . T h e t e chn i ca l des ign a n d per formance of u l tra fast c o m p u t e d tomography . Radiol. Clin. North Am., 32(3) :521-36, 1994. [51] J . A . Seibert . X - r a y p ro j e c t i on i m a g i n g . I n Imaging in Radiation Therapy. M e d i c a l P h y s i c s P u b l i s h i n g , M a d i s o n W I , 1998. [52] C . H . M c C o l l o u g h a n d F . E . Z i n k . C T numbers a n d e lectron density. I n Imaging in Radiation Therapy. M e d i c a l P h y s i c s P u b l i s h i n g , M a d i s o n W I , 1998. [53] D . J . Goodenough . T o m o g r a p h i c i m a g i n g . I n V a n M e t t e r R . L . B e u t e l J . , K u n d e l H . L . , edi tor , Handbook of Medical Imaging, Volume 1: Physics and Psychophysics. S P I E Press , 2000. [54] G E . G E M e d i c a l Systems: H i S p e e d C T / i ® T e c h n i c a l Reference M a n u a l . T e c h n i c a l report , 1999. [55] J . V . A t h e r t o n a n d W . H u d a . E n e r g y i m p a r t e d a n d effective doses i n c o m p u t e d tomography. Med. Phys., 23(5) :735-41, 1996. [56] W . H u d a , K . A . L i e b e r m a n , J . C h a n g , a n d M . L . Roskopf . P a t i e n t size a n d x -ray technique factors i n head c o m p u t e d t o m o g r a p h y e x a m i n a t i o n s . I. r a d i a t i o n doses. Med. Phys., 31(3) :588-94, 2004. [57] R . A . B r o o k s a n d G . D i C h i r o . S t a t i s t i c a l l i m i t a t i o n s i n x - r a y reconstruct ive tomography. Med. Phys., 3 (4) :237-40, 1976. [58] C . H . M c C o l l o u g h a n d F . E . Z i n k . Per f o rmance e v a l u a t i o n of a mul t i - s l i c e C T system. Med. Phys., 2 6 ( l l ) : 2 2 2 3 - 3 0 , 1999. [59] M . J . D a y a n d G . S te in . C h e m i c a l effects of i o n i z i n g r a d i a t i o n i n some gels. Nature, 166:146-7, 1950. [60] H . L . A n d r e w s , R . E . M u r p h y , a n d E . J . L e B r u n . G e l dos imeter for d e p t h dose measurements . Rev. Sci. Instrum., 28 :329-32, 1957. [61] J . C . Gore , Y . S. K a n g , a n d R . J . Schulz . T h e measurement of r a d i a t i o n dose d i s t r ibut i ons b y magnet i c resonance i m a g i n g . Magn. Reson. Imaging, 2:244, 1984. [62] L . E . O lsson , B . A . W e s t r i n , A . Fransson , a n d B . N o r d e l l . D i f fus i on of ferric ions i n agarose dosimeter gels. Phys. Med. Biol., 37 :2243-52 , 1992. [63] B . J . B a l c o m , T . J . Lees , A . R . S h a r p , N . S . K u l k a r n i , a n d G . S . W a g n e r . D i f -fusion i n F e ( I I / I I I ) r a d i a t i o n d o s i m e t r y gels measured b y magnet i c resonance imag ing . Phys. Med. Biol., 40 :1665-76, 1995. [64] C . B a l d o c k , P . J . H a r r i s , A . R . P i e r c y , a n d B . Hea ly . E x p e r i m e n t a l d e t e r m i -n a t i o n of the di f fus ion coefficient i n two-d imens ions i n ferrous su lphate gels us ing the f inite element m e t h o d . Australas. Phys. Eng. Sci. Med., 2 4 ( l ) : 1 9 - 3 0 , 2001. [65] Y . D e Deene a n d C . B a l d o c k , editors . Preliminary Proceedings of the 3rd Inter-national Conference on Radiation Therapy Gel Dosimetry. G h e n t U n i v e r s i t y , G h e n t , 2004. [66] M . J . M a r y a n s k i , R . J . Schulz , G . S. I b b o t t , J . C . Gatenby , J . X i e , D . H o r t o n , a n d J . C . Gore . M a g n e t i c resonance i m a g i n g of r a d i a t i o n dose d i s t r i b u t i o n s us ing p o l y m e r gel dosimeters. Phys. Med. Biol., 39 :1437-1455, 1994. [67] G . Ibbo t t , M . J . M a r y a n s k i , P . E a s t m a n , S. D . H o l c o m b , Y . Z h a n g , R . G . A v i s o n , M . Sanders , a n d J . C . G o r e . T h r e e - d i m e n s i o n a l v i s u a l i z a t i o n a n d measurement of c on formal dose d i s t r i b u t i o n s u s i n g magnet i c resonance i m a g -i n g of B A N G p o l y m e r gel dosimeters. Int. J. Radiat. Oncol. Biol. Phys., 38 (5) :1097-103, 1997. [68] Y . D e Deene, C . D e W a g t e r , B . V a n D u y s e , S. D e r y c k e , W . D e Neve , a n d E . A c h t e n . T h r e e - d i m e n s i o n a l d o s i m e t r y u s i n g p o l y m e r gel a n d magnet i c resonance i m a g i n g a p p l i e d to the ver i f i ca t i on of c on formal r a d i a t i o n t h e r a p y i n head-and-neck cancer. Radiother. Oncol, 48(3) :283-91 , 1998. [69] M . O l d h a m , I. B a u s t e r t , C . L o r d , T . A . D . S m i t h , M . M c J u r y , A . P . W a r -r i n g t o n , M . O . L e a c h , a n d S. W e b b . A n inves t iga t i on in to the d o s i m e t r y of a nine- f ie ld t o m o t h e r a p y i r r a d i a t i o n u s i n g B A N G - g e l dos imetry . Phys. Med. Biol, 43 :1113-32, 1998. [70] L . R i n t o u l , M . Lepage , a n d C . B a l d o c k . R a d i a t i o n dose d i s t r i b u t i o n i n p o l y m e r gels b y R a m a n spectroscopy. Appl. Sped., 5 7 ( l ) : 5 1 - 7 , 2003. [71] P . K e a l l a n d C . B a l d o c k . A theore t i ca l s t u d y of the rad io log i ca l propert ies a n d water equivalence of F r i c k e a n d p o l y m e r gels used for r a d i a t i o n dos imetry . Aus. Phys. Eng. Sci. Med., 22(3) :85-91 , 1999. [72] E . Pante l i s , A . K . K a r l i s , M . K o z i c k i , P . P a p a g i a n n i s , L . Sake l l i ou , a n d J . M . R o s i a k . P o l y m e r gel water equivalence a n d re lat ive energy response w i t h e m -phasis o n low p h o t o n energy d o s i m e t r y i n brachytherapy . Phys. Med. Biol., 49(15):3495-514, 2004. [73] J . Novotny , P . D v o r a k , V . Spevacek, J . T i n t e r a , T . Cechak , a n d R . L i s c a k . Q u a l i t y contro l of the stereotact ic rad iosurgery procedure w i t h the p o l y m e r -gel dosimetry. Radiother. and Oncol, 63(2) :223-30, 2002. [74] K . Vergote , Y . D e Deene, F . C l a u s , W . D e G e r s e m , B . V a n D u y s e , L . P a e l i n c k , E . A c h t e n , W . D e Neve , a n d C . D e W a g t e r . A p p l i c a t i o n of m o n o m e r / p o l y m e r gel dos imetry to s t u d y the effects of t issue inhomogeneit ies o n in tens i ty -m o d u l a t e d r a d i a t i o n t h e r a p y ( I M R T ) dose d i s t r i b u t i o n s . Rad. and Oncol., 67 ( l ) : 119 -28 , 2003. [75] Y . W a t a n a b e , R . M o o i j , G . M . P e r e r a , a n d M . J . M a r y a n s k i . Heterogenei ty p h a n t o m s for v i s u a l i z a t i o n of 3 D dose d i s t r i b u t i o n s b y M R I - b a s e d p o l y m e r gel dosimetry . Med. Phys., 31(5) :975-84, 2004. [76] A . J . Swal low. Radiation Chemistry: An Introduction. J o h n W i l e y a n d Sons, N e w Y o r k , 1973. [77] T - P . H s u a n d C . C o h e n . Observat ions o n the s t ruc ture of a p o l y a c r y l a m i d e gel f r o m electron micrographs . Polymer, 25 :1419-1423, 1984. [78] R . H . B o y d a n d P . J . P h i l l i p s . The Science of Polymer Molecules. C a m b r i d g e S o l i d State Science Series. C a m b r i d g e U n i v e r s i t y Press , C a m b r i d g e , 1996. [79] M . Lepage , A . K . W h i t t a k e r , L . R i n t o u l , S. A . B a c k , a n d C . B a l d o c k . T h e re la t ionship between r a d i a t i o n - i n d u c e d chemica l processes a n d transverse re-l a x a t i o n t imes i n p o l y m e r gel dosimeters . Phys. Med. Biol., 46(4) :1061-74, 2001. [80] A . J i rasek , C . D u z e n l i , C . A u d e t , a n d J . E l d r i d g e . C h a r a c t e r i z a t i o n of m o n o m e r / c r o s s l i n k e r c o n s u m p t i o n a n d p o l y m e r f o r m a t i o n observed i n F T -R a m a n spec t ra of i r r a d i a t e d p o l y a c r y l a m i d e gels. Phys. Med. Biol., 46 :151-65 , 2001. [81] C . B a l d o c k a n d Y . D e Deene, editors . Preliminary Proceedings of the 2nd International Conference on Radiation Therapy Gel Dosimetry. Q u e e n s l a n d Ins t i tu te of Technology, B r i s b a n e , 2001. [82] L . J . Schreiner a n d C . A u d e t , editors . Proceedings of the 1st International Workshop on Radiation Therapy Gel Dosimetry. C a n a d i a n O r g a n i z a t i o n of M e d i c a l P h y s i c i s t s a n d C a n a d i a n Col lege of P h y s i c i s t s i n M e d i c i n e , 1999. [83] M S D S . A c r y l a m i d e m a t e r i a l safety d a t a sheet. T e c h n i c a l repor t , S i g m a -A l d r i c h , 2004. [84] P . M . F o n g , D . C . K e i l , M . D . Does , a n d J . C . G o r e . P o l y m e r gels for magnet i c resonance i m a g i n g of r a d i a t i o n dose d i s t r i b u t i o n s at n o r m a l r o o m atmosphere . Phys. Med. Biol, 46 :3105-13 , 2001. [85] S. H e p w o r t h , M . O . L e a c h , a n d S . J . D o r a n . D y n a m i c s of p o l y m e r i z a t i o n i n p o l y a c r y l a m i d e gel ( P A G ) dosimeters : (II) m o d e l i n g oxygen di f fusion. Phys. Med. Biol, 44 :1875-84, 1999. [86] A . J i rasek a n d C . D u z e n l i . Effects of crossl inker f rac t i on i n p o l y m e r gel dos ime-ters us ing F T R a m a n spectroscopy. Phys. Med. Biol., 46 :1949-61 , 2001. [87] M . Lepage , A . K . W h i t t a k e r , L . R i n t o u l , a n d C . B a l d o c k . 1 3 C - N M R , X H -N M R a n d F T - R a m a n s t u d y of the r a d i a t i o n i n d u c e d modi f i cat ions i n r a d i a t i o n dos imetry p o l y m e r gels. J. Appl. Polym. Sci., 79 :1572-81, 2001. [88] K . Vergote , Y . D e Deene, E . V a n d e n Bussche , a n d C . D e W a g t e r . O n the r e l a t i o n between the s p a t i a l dose in tegr i ty a n d the t e m p o r a l i n s t a b i l i t y of p o l y m e r gel dosimeters. Phys. Med. Biol, 49 :4507-22, 2004. [89] V . Spevacek, J . N o v o t n y J r . , P . D v o r a k , J . Novotny , J . V y m a z a l , a n d T . Cechak . T e m p e r a t u r e dependence of po lymer -ge l dos imeter nuc lear m a g -net ic resonance response. Med. Phys., 2 8 ( l l ) : 2 3 7 0 - 7 8 , 2001. [90] G . J . Sa lomons , Y . S. P a r k , K . B . M c A u l e y , a n d L . J . Schreiner . T e m p e r a t u r e increases assoc iated w i t h p o l y m e r i z a t i o n of i r r a d i a t e d P A G dosimeters . Phys. Med. Biol., 47 :1435-48, 2002. [91] A . R . F a r a j o l l a h i . A n inves t igat i on in to the app l i ca t i ons of p o l y m e r gel dos ime-t r y i n radiotherapy . Med. Phys., 26(3):493, 1999. [92] A . R . F a r a j o l l a h i , D . E . B o n n e t t , A . J . Ratc l i f f e , R . J . A u k e t t , a n d J . A . M i l l s . A n invest igat ion into the use of p o l y m e r gel dos imetry i n low dose rate brachytherapy . Br. J. Radiol, 72 :1093-98, 1999. [93] C . B a l d o c k , A . G . Greener , N . C . B i l l i n g h a m , R . P . B u r f o r d , a n d S. F . K e e v i l . E n e r g y response a n d tissue equivalence of p o l y m e r gels for r a d i a t i o n d o s i m e t r y by M R I . Proc. Eur. soc. Magn. Resonance Med. Biol, 2:312, 1996. [94] J . N o v o t n y J r . , V . Spevacek, P . D v o r a k , J . Novotny , a n d T . C e c h a k . E n e r g y a n d dose rate dependence of B A N G - 2 p o l y m e r - g e l dosimeter . Med. Phys., 28 ( l l ) : 2 3 7 9 - 8 6 , 2001. [95] U . R a m m , M . D a m r a u , C . T h i l m a n n , a n d H . D . B o t t c h e r . 3 D heavy i o n dos imetry us ing magnet i c resonance i m a g i n g of p o l y m e r gels. I n L . J . Schreiner a n d C . A u d e t , editors , Proceedings of the 1st International Workshop on Radiation Gel Dosimetry, pages 181-183. C a n a d i a n Col lege of P h y s i c i s t s i n M e d i c i n e a n d C a n a d i a n O r g a n i z a t i o n of M e d i c a l P h y s i c i s t s , L e x i n g t o n K Y , 1999. [96] M . H i l t s , A J i rasek , C . A u d e t , a n d C . D u z e n l i . X - r a y C T p o l y m e r gel dos ime-t r y : app l i ca t i ons to stereotact ic rad iosurgery a n d p r o t o n therapy. Radiother. and Oncol, 56:S80, 2000. [97] A . J i r a s e k a n d C . D u z e n l i . R e l a t i v e effectiveness of p o l y a c r y l a m i d e gel dos ime-ters a p p l i e d to p r o t o n beams: F o u r i e r t r a n s f o r m R a m a n observat ions a n d t r a c k s t ruc ture ca lculat ions . Med. Phys., 29(4) :569-77, 2002. [98] C . A u d e t . NMR-dose Response Studies of the Gels Used for 3-D Radiation Dosimetry by Magnetic Resonance Imaging. P h D , M c G i l l U n i v e r s i t y , 1995. [99] C . B a l d o c k , M . Lepage , L . R i n t o u l , P . M u r r a y , a n d A . K . W h i t t a k e r . Invest i -ga t i on of p o l y m e r i z a t i o n of r a d i a t i o n d o s i m e t r y p o l y m e r gels. I n L . J . Schreiner a n d C . A u d e t , editors , Proceedings of the 1st International Workshop on Radi-ation Therapy Gel Dosimetry, pages 99 -105 . C a n a d i a n Col lege of P h y s i c i s t s i n M e d i c i n e a n d C a n a d i a n O r g a n i z a t i o n of M e d i c a l P h y s i c i s t s , E d m o n t o n , A B , 1999. [100] M . O . M c J u r y , M . O . L e a c h , a n d S. W e b b . D y n a m i c s of p o l y m e r i z a t i o n i n p o l y a c r y l a m i d e gel ( P A G ) dosimeters : (I) ageing a n d l o n g t e r m s tab i l i ty . Phys. Med. Biol, 44 :1863-73 , 1999. [101] Y . D e Deene, P . Hanse laer , C . D e W a g t e r , E . A c h t e n , a n d W . D e Neve . A n inves t igat i on of the chemica l s t a b i l i t y of a m o n o m e r / p o l y m e r gel dosimeter . Phys. Med. Biol, 45(4) :859-78, 2000. [102] M . Lepage , A . K . W h i t t a k e r , L . R i n t o u l , S. A . B a c k , a n d C . B a l d o c k . M o d -e l l ing of p o s t - i r r a d i a t i o n events i n p o l y m e r gel dosimeters . Phys. Med. Biol., 4 6 ( l l ) : 2 8 2 7 - 3 9 , 2001. [103] Y . D e Deene, A . V e n n i n g , C . H u r l e y , B . J . Hea ly , a n d C . B a l d o c k . Dose -response s t a b i l i t y a n d in tegr i ty of the dose d i s t r i b u t i o n of various p o l y m e r gel dosimeters. Phys. Med. Biol., 47(14) :2459-70, 2002. [104] Y . D e Deene, N . R e y n a e r t , a n d C . D e W a g t e r . O n the accuracy of m o n o m e r / p o l y m e r gel dos imetry i n the p r o x i m i t y of a high-dose-rate 1 9 2 I r source. Phys. Med. Biol., 4 6 ( l l ) : 2 8 0 1 - 2 5 , 2001. [105] K . Vergote , Y . D e Deene, W . D u t h o y , W . D e G e r s e m , W . D e Neve , E . A c h t e n , a n d C . D e Wagter . V a l i d a t i o n a n d a p p l i c a t i o n of p o l y m e r gel dos imetry for the dose ver i f i cat ion of a n i n t e n s i t y - m o d u l a t e d arc t h e r a p y ( I M A T ) t reatment . Phys. Med. Biol., 49(2) :287-305, 2004. [106] C . Hur l ey , Y . D e Deene, R . M e d e r , J . M . P o p e , a n d C . B a l d o c k . T h e effect of water molecu lar self -dif fusion o n q u a n t i t a t i v e h igh - reso lu t i on M R I p o l y m e r gel dosimetry . Phys. Med. Biol, 48(18) :3043-58, 2003. [107] A . M . F u x m a n , K . B . M c A u l e y , a n d L . J . Schreiner . M o d e l i n g of f ree-radical c ross l ink ing c o p o l y m e r i z a t i o n of a c r y l a m i d e a n d N , N ' -methy leneb is (acry lamide ) for r a d i a t i o n dos imetry . Macromol. Theory Simul., 12(9):647-62, 2003. [108] K . B . M c A u l e y , A . M . F u x m a n , a n d L . J . Schreiner . M o d e l i n g chemica l a n d p h y s i c a l p h e n o m e n a i n p o l y a c r y l a m i d e gel dos imeters . I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004: 3rd Interna-tional Conference on Radiotherapy Gel Dosimetry, pages 149-152. G h e n t U n i -versity, G h e n t , B e l g i u m , 2004. [109] E . P a p p a s , T . M a r i s , A . Ange lopou los , M . P a p a r i g o p o u l o u , L . Sake l l i ou , P . Sandi los , S. V o y i a t z i , a n d L . V l a c h o s . A new p o l y m e r gel for magnet i c resonance i m a g i n g ( M R I ) r a d i a t i o n dos imetry . Phys. Med. Biol., 44 :2677-84, 1999. [110] P . K i p o u r o s , E . P a p p a s , P . B a r a s , D . H a t z i p a n a y o t i , P . K a r a i s k o s , L . Sake l l i ou , P . Sandi los , a n d I. Seimenis . W i d e d y n a m i c dose range of V I P A R p o l y m e r gel dos imetry . Phys. Med. Biol, 46(8) :2143-59, 2001. [ I l l ] M . Lepage , P . M . Jayasakera , S. A . J . B a c k , a n d C . B a l d o c k . Dose reso lut i on o p t i m i z a t i o n of p o l y m e r gel dosimeters u s i n g different monomers . Phys. Med. Biol, 46 :2665-80, 2001. [112] H . G u s t a v s s o n , S. A . J . B a c k , M . Lepage , L . R i n t o u l , a n d C . B a l d o c k . Deve l op -ment a n d o p t i m i z a t i o n of a 2 -hydroxye thy lac ry la te M R I p o l y m e r gel dos ime-ter . Phys. Med. Biol, 49(2) :227-241, 2004. [113] Y . D e Deene, C . H u r l e y , A . V e n n i n g , K . Vergote , M . M a t h e r , B . J . Hea ly , a n d C . B a l d o c k . A basic s t u d y of some n o r m o x i c p o l y m e r gel dosimeters . Phys. Med. Biol, 47(19) :3441-63, 2002. [114] M . H . Z a h m a t k e s h , R . K o u s a r i , A k h l a g h p o u r S., a n d S. A . B a g h e r i . M R I gel dos imetry w i t h metac ry l i c a c i d , ge la t in , ascorbic a c i d , h y d r o q u i n o n e a n d copper i n agarose ( M A G I C A ) gel . I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL200J,.: 3rd International Conference on Radiotherapy Gel Dosimetry, pages 134-139. G h e n t U n i v e r s i t y , G h e n t , B e l -g i u m , 2004. [115] J . H r b a c e k , V . Spevacek, J . N o v o t n y J r . , a n d T . Cechak . A comparat ive s t u d y of four p o l y m e r gel dosimeters. I n Y . D e Deene a n d C . B a l d o c k , editors , Pre-liminary proceedings of DOSGEL2004 : 3rd International Conference on Ra-diotherapy Gel Dosimetry, pages 140-144. G h e n t U n i v e r s i t y , G h e n t , B e l g i u m , 2004. [116] A . V e n n i n g , S. B r i n d h a , B . H i l l , a n d C . B a l d o c k . P r e l i m i n a r y s t u d y of a nor -mox i c P A G gel dosimeter w i t h t e t rak i s (hydroxymethy l ) p h y o s p h o n i u m chlo -r ide as a n ant i - ox idant . I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004 : 3rd International Conference on Radiotherapy Gel Dosimetry, pages 145-148. G h e n t U n i v e r s i t y , G h e n t , B e l g i u m , 2004. [117] E . V a n d e n Bussche , Y . D e Deene , K . Vergote , a n d C . D e W a g t e r . A l t e r -nat ive ge l l ing agents for n o r m o x i c gels: a s t a b i l i t y s tudy. I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004 : 3rd Inter-national Conference on Radiotherapy Gel Dosimetry, pages 158-161 . G h e n t U n i v e r s i t y , G h e n t , B e l g i u m , 2004. [118] S. B r i n d h a , A . V e n n i n g , B . H i l l , a n d C . B a l d o c k . E x p e r i m e n t a l s t u d y of a t t e n u a t i o n propert ies of n o r m o x i c p o l y m e r gel dosimeters . Phys. Med. Biol., 4 9 : N 3 5 3 - N 3 6 1 , 2004. [119] C . B a l d o c k , M . Lepage , S. A . B a c k , P . J . M u r r y , P . M . Jayasekera , D . P o r t e r , a n d T . K r o n . Dose reso lut i on i n r a d i o t h e r a p y p o l y m e r gel dos imetry : effect of echo spac ing i n M R I pulse sequence. Phys. Med. Biol., 46(2) :449-60, 2001. [120] Y . D e Deene a n d C . B a l d o c k . O p t i m i z a t i o n of m u l t i p l e spin-echo sequences for 3 D p o l y m e r gel dos imetry . Phys. Med. Biol., 47(17) :3117-41, 2002. [121] A . B a n k a m p a n d L . R . Schad . C o m p a r i s o n of T S E , T G S E , a n d C P M G m e a -surement techniques for M R p o l y m e r gel dos imetry . Magn. Reson. Imag., 21 (8) :929-39, 2003. [122] P . B a r a s , I. Seimenis , P K i p o u r o s , P . P a p a g i a n n i s , A . Ange lopou los , L . S a k e l -l i o u , E . P a p p a s , D . B a l t a s , P . K a r a i s k o s , P . Sandios , a n d L . V l a c h o s . P o l y m e r gel dos imetry us ing a three -d imens i ona l M R I a c q u i s i t i o n technique . Med. Phys., 29(11) :2506-16, 2002. [123] P . M a g n u s s o n a n d L . E . O l s son . Image analys is methods for assessing levels of image p lane n o n u n i f o r m i t y a n d stochast ic noise i n a magnet i c resonance image of a homogeneous p h a n t o m . Med Phys, 27(8) :1980-94, 2000. [124] D . A . L o w , J . M a r k m a n , J . F . Dempsey , S. M u t i c , M . O l d h a m , R . Venkatesan , E . M . Haacke , a n d J . A . P u r d y . No ise i n p o l y m e r gel measurements us ing M R I . Med. Phys., 27(8) :1814-17, 2000. [125] Y . W a t a n a b e , G . M . P e r e r a , a n d R . B . M o o i j . Image d i s t o r t i o n i n M R I - b a s e d p o l y m e r gel dos imetry of G a m m a K n i f e stereotact ic rad iosurgery systems. Med. Phys., 29(5) :797-802, 2002. [126] W . D u t h o y , W . D e G e r s e m , K . Vergote , M . Coghe , T . B o t e r b e r g , Y . D e Deene, C . D e W a g t e r , S. V a n B e l l e , a n d W . D e Neve . W h o l e a b d o m i n o p e l v i c rad io therapy ( W A P R T ) u s i n g i n t e n s i t y - m o d u l a t e d arc t h e r a p y ( I M A T ) : f irst c l i n i c a l experience. Int. J. Rad. Oncol. Biol. Phys., 57(4) :1019-1032, 2003. [127] A . B e r g , A . E r t l , a n d E . M o s e r . H i g h reso lut i on p o l y m e r gel dos imetry by parameter selective M R - m i c r o i m a g i n g o n a whole b o d y scanner at 3 T . Med. Phys., 28(5) :833-43, 2001. [128] Y . X u , C . S. W u u , a n d M . J . M a r y a n s k i . D e t e r m i n i n g o p t i m a l gel sens i t iv i ty i n o p t i c a l C T scanning of gel dosimeters . Med. Phys., 30(8) :2257-63, 2003. [129] K . T . S. I s lam, J . F . Dempsey , M . K . R a n a d e , M . J . M a r y a n s k i , a n d D . A . L o w . I n i t i a l eva luat i on of c o m m e r c i a l o p t i c a l C T - b a s e d 3 D gel dos imeter . Med. Phys., 30(8) :2159-68, 2003. [130] J . G . W o l o d z k o , C . M a r s d e n , a n d A . A p p l e b y . C C D i m a g i n g for o p t i c a l t o -m o g r a p h y of gel r a d i a t i o n dosimeters . Med. Phys., 26(11) :2508-13, 1999. [131] S. J . D o r a n , K . K . K o e r k a m p , M . A . B e r o , P . Jenneson , E . J . M o r t o n , a n d W . B . G i l b o y . A C C D - b a s e d o p t i c a l C T scanner for h igh - reso lu t i on 3 D i m a g i n g of r a d i a t i o n dose d i s t r i b u t i o n s : equipment speci f icat ions, o p t i c a l s imulat ions a n d p r e l i m i n a r y results . Phys. Med. Biol, 46 :3191-3213, 2001. [132] M . O l d h a m , J . H . Siewerdsen, S. K u m a r , J . W o n g , a n d D . A . Jaffray. O p t i c a l -C T ge l -dos imetry I: B a s i c invest igat ions . Med. Phys., 28(7) :1436-45, 2003. [133] M . O l d h a m a n d L . K i m . O p t i c a l - C T ge l -dos imetry II : O p t i c a l art i facts a n d geometr ica l d i s t o r t i o n . Med. Phys., 31:1093-104, 2004. [134] M . O l d h a m , G . G l u c k m a n , a n d L . K i m . 3 D ver i f i ca t i on of a prostate I M R T treatment b y p o l y m e r ge l -dos imetry a n d o p t i c a l - C T s cann ing . I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004- 3rd Inter-national Conference on Radiotherapy Gel Dosimetry, pages 284-287. G h e n t Un ivers i ty , G h e n t , B e l g i u m , 2004. [135] C . W u u , Y . X u , a n d M . M a r y a n s k i . 3 D dose ver i f i ca t i on for I M R T us ing op -t i c a l ct based p o l y m e r gel dos imetry . I n Y . D e Deene a n d C B a l d o c k , editors , Preliminary proceedings of DOSGEL2004: 3rd International Conference on Radiotherapy Gel Dosimetry, pages 288-291 . G h e n t U n i v e r s i t y , G h e n t , B e l -g i u m , 2004. [136] S . J . D o r a n , N . K r s t a j i c , J . A d a m o v i c s , a n d P . M . Jenneson . O p t i c a l C T scanning of p r e s a g e ™ p o l y u r e t h a n e samples w i t h a C C D - b a s e d readout sys-t e m . I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004: 3rd International Conference on Radiotherapy Gel Dosime-try, pages 231-234. G h e n t U n i v e r s i t y , G h e n t , B e l g i u m , 2004. [137] K . J . J o r d a n . A d v a n c e s i n o p t i c a l C T s cann ing for gel dos imetry . I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOS-GEL2004- 3rd International Conference on Radiotherapy Gel Dosimetry, pages 98-104. G h e n t U n i v e r s i t y , G h e n t , B e l g i u m , 2004. [138] J . T r a p p , G . M i c h a e l , a n d C . B a l d o c k . T h e o r e t i c a l considerat ions of scan p a -rameters appropr ia te for C T i m a g i n g of p o l y m e r gel dosimeters . I n C . B a l d o c k a n d Y . D e Deene, editors , Proc. 2nd International Con], on Radiotherapy Gel Dosimetry, B r i s b a n e , A U S , 2001. Q u e e n s l a n d U n i v e r s i t y of Technology. [139] J . T r a p p , G . M i c h a e l , a n d C . B a l d o c k . U n c e r t a i n t y ca l cu lat ions de te rmined f r o m c a l i b r a t i o n d a t a for C T i m a g i n g of p o l y m e r gel dosimeters . I n C . B a l d o c k a n d Y . D e Deene, editors , Proc. 2nd International Conf. on Radiotherapy Gel Dosimetry, B r i s b a n e , A U S , 2001. Q u e e n s l a n d U n i v e r s i t y of Technology. [140] J . V . T r a p p , G . M i c h a e l , Y . de Deene, a n d C . B a l d o c k . A t t e n u a t i o n of d i a g -nost ic energy photons b y p o l y m e r gel dosimeters. Phys. Med. Biol., 47(23): 4247-58 , 2002. [141] P . M . Jenneson, E . C . A t k i n s o n , P . W a i , a n d S . J . D o r a n . O p t i m i s a t i o n of x - r a y m i c r o - t o m o g r a p h y for the low-dose analys is of h igh ly -dosed gels. I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOS-GEL2004: 3rd International Conference on Radiotherapy Gel Dosimetry, pages 248-251 . G h e n t U n i v e r s i t y , G h e n t , B e l g i u m , 2004. [142] B . H i l l , A . V e n n i n g , a n d C . B a l d o c k . Use of n o r m o x i c p o l y m e r gel dosimeters for measur ing d iagnost i c doses o n C T scanners. I n Y . D e Deene a n d C . B a l -dock, editors , Preliminary proceedings of DOSGEL2004-' 3rd International Conference on Radiotherapy Gel Dosimetry, pages 214-217 . G h e n t U n i v e r -sity, G h e n t , B e l g i u m , 2004. [143] M . H i l t s a n d C . D u z e n l i . Image f i l t er ing for i m p r o v e d dose reso lut i on i n C T p o l y m e r gel dosimetry . Med. Phys., 3 1 ( l ) : 3 9 - 4 9 , 2004. [144] M . H i l t s , A . J i rasek , a n d C . D u z e n l i . Effects of gel c o m p o s i t i o n o n the r a d i -a t i o n induced dens i ty change i n P A G p o l y m e r gel dosimeters : a m o d e l a n d exper imenta l invest igat ions . Phys. Med. Biol, 49 :2477-90, 2004. [145] M . H i l t s , A . J i rasek , a n d C . D u z e n l i . T e c h n i c a l considerat ions for i m p l e m e n -t a t i o n of x - r a y C T p o l y m e r gel dos imetry . Phys. Med. Biol., 50:1727-1745, 2005. [146] M . H i l t s , A . J i rasek , a n d C . D u z e n l i . T h e response of P A G dens i ty to dose: a m o d e l a n d e x p e r i m e n t a l invest igat ions . I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004: 3rd International Confer-ence on Radiotherapy Gel Dosimetry, pages 153-157. G h e n t Un ivers i ty , G h e n t , B e l g i u m , 2004. [147] M . H i l t s a n d C . D u z e n l i . Image noise i n x - r a y C T p o l y m e r gel dos ime-try . I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004: 3rd International Conference on Radiotherapy Gel Dosime-try, pages 243-247 . G h e n t U n i v e r s i t y , G h e n t , B e l g i u m , 2004. [148] M . L . M a t h e r , Y . D e Deene, A . K . W h i t t a k e r , G . S i m o n , R . R u t g e r s , a n d C . B a l d o c k . Invest igat ion of u l t rason i c propert ies of P A G a n d M A G I C p o l y m e r gel dosimeters . Phys. Med. Biol., 47 :4397-409, 2002. [149] M . L . M a t h e r , P . H . C h a r l e s , a n d C . B a l d o c k . M e a s u r e m e n t of u l t rason i c a t t e n u a t i o n coefficient i n p o l y m e r gel dosimeters . Phys. Med. Biol, 48(20): N 2 6 9 - N 2 7 5 , 2003. [150] M . L . M a t h e r a n d C . B a l d o c k . U l t r a s o u n d t o m o g r a p h i c i m a g i n g of r a d i a t i o n dose d i s t r i b u t i o n s i n p o l y m e r gel dosimeters : P r e l i m i n a r y study. Med. Phys., 30(8) :2140-48, 2003. [151] M . Lepage , P . S. Tof ts , S. A . B a c k , P . M . Jayasekera , a n d C . B a l d o c k . S i m p l e methods for the correc t i on of T 2 m a p s of p h a n t o m s . Magn. Reson. Med., 46 (6): 1123-9, 2001. [152] L . E . O l s son , S. A . J . B a c h , P . M a g n u s s o n , a n d P . H a r a l d s s o n . 3 D dos imetry us ing gels a n d M R I . I n Imaging in Radiation Therapy, pages 475-04 . M e d i c a l P h y s i c s P u b l i s h i n g , W I , 1998. [153] M . M c J u r y , M . O l d h a m , V . P . Cosgrove, P . S. M u r p h y , S. D o r a n , M . O. L e a c h , a n d S. W e b b . R a d i a t i o n d o s i m e t r y us ing p o l y m e r gels: methods a n d app l i ca t i ons . Brit. J. Radiol., 73 :919-29, 2000. [154] P . A . L o v e , P . M . E v a n s , M . O . L e a c h , a n d S. W e b b . P o l y m e r gel measurement of dose homogenei ty i n the breast : c o m p a r i n g M L C in tens i ty m o d u l a t i o n w i t h s t a n d a r d wedged del ivery. Phys. Med. Biol, 48(8) :1065-1074, 2003. [155] D . A . L o w , J . F . Dempsey , R . Venkatesan , S. M u l t i c , J . M a r k m a n , E . M . H a a c k e , a n d J . A . P u r d y . E v a l u a t i o n of p o l y m e r gels a n d M R I as a 3 -D dosimeter for i n t e n s i t y - m o d u l a t e d r a d i a t i o n therapy. Med. Phys., 26(8): 1 5 4 2 -51, 1999. [156] Y . D e Deene. G e l dos imetry for the dose ver i f i cat ion of in tens i ty m o d u l a t e d rad io therapy t reatments . Z. Med. Phys., 12(2) :77-88, 2002. [157] H . G u s t a v s s o n , A . K a r l s s o n , S. A . J . B a c k , L . E . O l s son , P . H a r a l d s s o n , P . E n -gs t rom, a n d H . N y s t r o m . M A G I C - t y p e p o l y m e r gel for three -d imens i ona l dos imetry : I n t e n s i t y - m o d u l a t e d r a d i a t i o n therapy ver i f i cat ion . Med. Phys., 30(6) :1264-1271, 2003. [158] J . N o v o t n y J r . , J . Novotny , V . Spevacek, P . D v o r a k , T . C e c h a k , R . L i s c a k , G . B r o z e k , J . T i n t e r a , a n d J . V y m a z a l . A p p l i c a t i o n of p o l y m e r gel d o s i m e t r y i n g a m m a kni fe radiosurgery. J. Neurosurg., 97(5 S u p p l ) : 5 5 6 - 6 2 , 2002. [159] J . N o v o t n y J r . , J . Novotny , V . Spevacek, P . D v o r a k , T . Cechak , R . L i s c a k , G . Brozek , J . T i n t e r a , a n d J . V y m a z a l . E v a l u a t i o n of geometr ic a n d dos imet -r i c inaccuracies of stereotact ic i r r a d i a t i o n i n the rat b r a i n . Stereotact Funct Neurosurg, 79(2) :57-74, 2002. [160] J . N o v o t n y J r . , P . D v o r a k , V . Spevacek, J . T i n t e r a , J . Novotny , a n d T . Cechak . M e d i c a l a p p l i c a t i o n of 3 -D p o l y m e r gel dosemeter eva luated b y nuclear m a g -net ic resonance. Radiat Prot Dosimetry, 101(1-4) :399-402, 2002. [161] E . P a p p a s , I. Seimenis , A . Ange lopou los , P . Georgo l opou lou , M . K a m a r i o t a k i -P a p a r i g o p o u l o u , T . M a r i s , L . Sake l l i ou , P . Sandi l os , a n d L . V l a c h o s . N a r r o w stereotactic b e a m profile measurements u s i n g N - v i n y l p y r r o l i d o n e based p o l y -mer gels a n d magnet i c resonance i m a g i n g . Phys. Med. Biol., 46 :783-97, 2001. [162] G . Grebe , M . P faender , M . R o l l , L . L u e d e m a n n , a n d R . E . W u r m . D y n a m i c arc rad iosurgery a n d rad io therapy : c o m m i s s i o n i n g a n d ver i f i cat ion of dose d i s t r ibut i ons . Int. J. Rad. Oncol. Biol. Phys., 49(5) :1451-1460, 2001. [163] V . P . Cosgrove, P . S. M u r p h y , M . M c J u r y , E . J . A d a m s , A . W a r r i n g t o n , M . O . L e a c h , a n d S. W e b b . T h e r e p r o d u c i b i l i t y of p o l y a c r y l a m i d e gel dos imetry app l i ed to stereotact ic con formal rad io therapy . Phys. Med. Biol., 45 (5 ) :1195-1210, 2000. [164] P . K i p o u r o s , P . P a p a g i a n n i s , L . Sake l l i ou , P . K a r a i s k o s , P . Sandi los , P . B a r a s , I. Seimenis , a n d M . K o z i c k i . 3 D dose ver i f i ca t i on i n 1-192 H D R prostate monotherapy us ing p o l y m e r gels a n d M R I . Med. Phys., 30(8) :2031-39, 2003. [165] M . O . M c J u r y , P . D . T a p p e r , V . P . Cosgrove, P . S. M u r p h y , S. G r i f f i n , M . O . L e a c h , S. W e b b , a n d M . O l d h a m . E x p e r i m e n t a l 3 D d o s i m e t r y a r o u n d a h i g h -dose rate c l i n i c a l 1 9 2 I r source us ing a p o l y a c r y l a m i d e gel ( P A G ) dosimeter . Phys. Med. Biol, 44 :1863-73, 1999. [166] P . P a p a g i a n n i s , E . P a p p a s , P . K i p o u r o s , A . Ange l opou los , L . Sake l l i ou , P . B a r a s , P . K a r a i s k o s , I. Seimenis , P . Sandi los , a n d D . B a l t a s . D o s i m e t r y close to a n 1 9 2 I r H D R source us ing N - v i n y l p y r r o l i d o n e based p o l y m e r gels a n d magnet i c resonance imag ing . Med. Phys., 28(7) :1416-26, 2001. [167] C . W u u , P . Scruff, M . J . M a r y a n s k i , T . L i u , S. B o r z i l l a r y , a n d J . Weinberger . D o s i m e t r y s t u d y of Re-188 l i q u i d b a l l o o n for in t ravascu lar b r a c h y t h e r a p y us -i n g p o l y m e r gel dosimeters a n d laser -beam o p t i c a l C T scanner. Med. Phys., 30(2) :132-7, 2003. [168] M . F . C h a n , A . Y . F u n g , Y . C . H u , C . S. C h u i , H . A m o l s , M . Za ider , a n d D . A b r a m s o n . T h e measurement of three d i m e n s i o n a l dose d i s t r i b u t i o n of a r u t h e n i u m - 1 0 6 o p h t h a l m o l o g i c a l a p p l i c a t o r us ing magnet i c resonance i m a g i n g of B A N G p o l y m e r gels. J Biochem. Biophs. Meth., 1:105-116, 2001. [169] U . O . H a f e l i , W . R . R o b e r t s , D . S. M e i e r , J . P . C i e z k i , G . J . P a u e r , E . J . Lee , a n d M . S. We inhous . D o s i m e t r y of a W - 1 8 8 / R e - 1 8 8 b e t a l ine source for endovascular brachytherapy . Med. Phys., 27(4) :668-675, 2000. [170] U . R a m m , U . W e b e r , M . B o c k , M . K r a m e r , A . B a n k a m p , M . D a m r a u , C . T h i l m a n n , H . D . B o t t c h e r , L . R . S c h a d , a n d G . K r a f t . T h r e e - d i m e n s i o n a l B A N G ™ gel dos imetry i n c o n f o r m a l c a r b o n i o n radiotherapy . Phys. Med. Biol., 4 5 : N 9 5 - N 1 0 2 , 2000. [171] J . Heufe lder , S. Stiefel , M . P faender , L . L u d e m a n n , G . G r e b e , a n d J . Heese. U s e of B A N G p o l y m e r gel for dose measurements i n a 68 M e V p r o t o n b e a m . Med. Phys., 30(6) :1235-40, 2003. [172] J . U u s i - S i m o l a , S. Savo la inen , A . K a n g a s m a k i , S. H e i k k i n e n , J . P e r k i o , U . A b o R a m a d a n , T . Seppala , J . K a r i l a , T . Seren, P . K o t i l u o t o , P . S o r v a r i , a n d I. A u -ter inen . S t u d y of the re lat ive dose-response of B A N G - 3 p o l y m e r gel dosimeters i n e p i t h e r m a l n e u t r o n i r r a d i a t i o n . Phys. Med. Biol., 48(17) :2895-906, 2003. [173] C . W o j n e c k i a n d S. G r e e n . A c o m p u t a t i o n a l s t u d y in to the use of p o l y a c r y -l a m i d e gel a n d A - 1 5 0 p las t i c as b r a i n t issue subst i tutes for b o r o n n e u t r o n capture therapy. Phys. Med. Biol., 46(5) :1399-1405, 2001. [174] A . R . F a r a j o l l a h i , D . E . B o n n e t t , D . T a t t a m , a n d S. G r e e n . T h e p o t e n t i a l use of p o l y m e r gel dos imetry i n b o r o n n e u t r o n capture therapy. Phys. Med. Biol., 4 5 : N 9 - N 1 4 , 2000. [175] L . B o u c h a r d a n d M . J . B r o n s k i l l . M a g n e t i c resonance i m a g i n g of t h e r m a l coagulat i on effects i n a p h a n t o m for c a l i b r a t i n g t h e r m a l t h e r a p y devices. Med. Phys., 27(5) :1141-45, 2000. [176] K . G i l l f o r d , J . H o r t o n , T . Steger, M . P . H e a r d , E . J a c k s o n , a n d G . Ibbo t t . Ver i f i ca t i on of M o n t e C a r l o ca l cu la t i ons a r o u n d a F l e t c h e r Su i t Delc los ovo id w i t h normox i c p o l y m e r gel dos imetry . I n Y . D e Deene a n d C B a l d o c k , e d i -tors , Preliminary proceedings of DOSGEL2004' 3rd International Conference on Radiotherapy Gel Dosimetry, pages 207-210. G h e n t U n i v e r s i t y , G h e n t , B e l -g i u m , 2004. [177] M . P . H e a r d a n d G . Ibbo t t . M e a s u r e m e n t of b r a c h y t h e r a p y sources us ing M A G I C gel. I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004 : 3rd International Conference on Radiotherapy Gel Dosime-try, pages 211-213. G h e n t U n i v e r s i t y , G h e n t , B e l g i u m , 2004. [178] S . G . Scheib, Y . Schenkel , a n d S. G i a n o l i n i . A b s o l u t e dose ver i f i ca t i on i n s m a l l p h o t o n fields us ing B A N G ™ gel. I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004- 3rd International Conference on Radiotherapy Gel Dosimetry, pages 218-221 . G h e n t U n i v e r s i t y , G h e n t , B e l -g i u m , 2004. [179] C . Bayreder , A . B e r g , D . G e o r g , a n d E . M o s e r . H i g h reso lut i on dos imetr i c i m a g i n g of p o l y m e r gels prepared under n o r m o x i c cond i t i ons : s tereotact ic r a -d i a t i o n w i t h a 16 m m c o l l i m a t o r . I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004'- 3rd International Conference on Radiotherapy Gel Dosimetry, pages 263-266. G h e n t U n i v e r s i t y , G h e n t , B e l -g i u m , 2004. [180] W . D u t h o y , K . Vergote , W . D e G e r s e m , Y . D e Deene, C D e W a g t e r , a n d W . D e Neve . C l i n i c a l v a l i d a t i o n of in tens i ty m o d u l a t e d arc t h e r a p y ( I M A T ) by means of p o l y m e r gel dos imetry . I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004'- 3rd International Conference on Radiotherapy Gel Dosimetry, pages 267-270. G h e n t U n i v e r s i t y , G h e n t , B e l -g i u m , 2004. [181] J . T r a p p , A . W a r r i n g t o n , M . P a r t r i d g e , A . P h i l i p s , M . O . L e a c h , a n d S. W e b b . 3 D measurement of absolute r a d i a t i o n dose i n g r i d therapy. I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004-' 3rd Inter-national Conference on Radiotherapy Gel Dosimetry, pages 271-274. G h e n t Un ivers i ty , G h e n t , B e l g i u m , 2004. [182] A . K a r l s s o n , B . Smulders , H . G u s t a v s s o n , a n d S. A . J . B a c h . I M R T prostate dos imetry us ing a n o r m o x i c p o l y m e r gel a n d M R I . I n Y . D e Deene a n d C . B a l -dock, editors , Preliminary proceedings of DOSGEL2004'- 3rd International Conference on Radiotherapy Gel Dosimetry, pages 275-278. G h e n t U n i v e r -sity, G h e n t , B e l g i u m , 2004. [183] J . N o v o t n y J r . , V . Spevacek, P . D v o r a k , J . H r b a c e k , J . Novotny , D . T l a c h a -cova, M . S c h m i t t , J . V y m a z a l , J . T i n t e r a , a n d T . Cechak . A p p l i c a t i o n of p o l y -mer gel d o s i m e t r y i n stereotact ic radiosurgery. I n Y . D e Deene a n d C . B a l d o c k , editors , Preliminary proceedings of DOSGEL2004: 3rd International Confer-ence on Radiotherapy Gel Dosimetry, pages 279-283 . G h e n t U n i v e r s i t y , G h e n t , B e l g i u m , 2004. [184] J . H . H u b b e l l . P h o t o n cross sections, a t t e n u a t i o n coefficients, a n d energy a b -sorpt i on coefficients f r o m 10 k e V to 100 G e V . T e c h n i c a l report , U S D e p a r t -ment of C o m m e r c e - N a t i o n a l B u r e a u of S t a n d a r d s , 1969. [185] M . H i l t s , C . A u d e t , C . D u z e n l i , a n d J . R o b a r . P o l y m e r gel dos imetry u s i n g x - ray c o m p u t e d t omography : feas ib i l i ty a n d p o t e n t i a l a p p l i c a t i o n to stereo-tac t i c radiosurgery. I n L . J . Schreiner a n d C . A u d e t , editors , Proceedings of the 1st international workshop on radiation gel dosimetry, pages 190-192. C a n a -d i a n Col lege of P h y s i c i s t s i n M e d i c i n e a n d C a n a d i a n O r g a n i z a t i o n of M e d i c a l P h y s i c i s t s , E d m o n t o n , A B , 1999. [186] I n t e r n a t i o n a l C o m m i s s i o n on R a d i a t i o n U n i t s a n d M e a s u r e m e n t ( I C R U ) . I C R U R e p o r t N o . 42: Use of computers i n e x t e r n a l b e a m r a d i o t h e r a p y proce-dures w i t h h i g h energy photons a n d electrons. T e c h n i c a l repor t , I C R U P u b -l i cat ions , 1987. [187] M . O l d h a m , M . M c J u r y , I. B a u s t e r t , M . O . L e a c h , a n d S. W e b b . I m p r o v i n g c a l i b r a t i o n accuracy i n gel dos imetry . Phys. Med. Biol., 43 :2709-20, 1998. [188] C . B a l d o c k , P . M u r r y , a n d T . K r o n . U n c e r t a i n t y analys is i n p o l y m e r gel dos imetry . Phys. Med. Biol., 4 4 : N 2 4 3 - 6 , 1999. [189] I S O . G u i d e to the expression of u n c e r t a i n t y i n measurement . T e c h n i c a l report , I S O , 1995. [190] J . V . T r a p p , G . M i c h a e l , P . M . E v a n s , C . B a l d o c k , M . O . L e a c h , a n d S. W e b b . Dose reso lut ion i n gel dos imetry : effect of u n c e r t a i n t y i n the c a l i b r a t i o n func -t i o n . Phys. Med. Biol, 4 9 : N 1 3 9 - N 1 4 6 , 2004. [191] A . J i rasek . Fourier Transform Raman Spectroscopy of Polyacrylamide Gels for use in Radiation Dosimetry. P h D , U n i v e r s i t y of B r i t i s h C o l u m b i a , 2002. [192] W T W . O x i 330i /340z opera t ing m a n u a l . T e c h n i c a l r epor t , W T W Wissenschaf t l i ch -Technische W e r k s t a t t e n G m b H , 2000. [193] D . E . B o n n e t t , A . J . F a r a j o l l a h i , a n d P . F . H a r d i n g . A n inves t iga t i on i n t o the w a l l mater ia l s for p o l y m e r gel phantoms . I n L . J . Schreiner a n d C . A u d e t , editors , Proceedings of the 1st International Workshop on Radiation Therapy Gel Dosimetry, pages 142-144. C a n a d i a n Col lege of P h y s i c i s t s i n M e d i c i n e a n d C a n a d i a n O r g a n i z a t i o n of M e d i c a l P h y s i c i s t s , E d m o n t o n , A B , 1999. [194] W h e a t o n Scienti f ic . T e c h n i c a l D a t a Sheet: H i g h D e n s i t y Po lye thy lene ( H D P E ) . T e c h n i c a l report , W h e a t o n Scienti f ic P r o d u c t s , 2003. [195] C . J . K a r z m a r k , S. N u n a n , a n d E . T a n a b e . Medical Electron Accelerators. M c G r a w H i l l Inc. , N Y , 1992. [196] C . G e l f i a n d P . G . R i g h e t t i . P o l y m e r i z a t i o n k inet i cs of p o l y a c r y l a m i d e gels I. Effects of different crossl inkers. Electrophoresis, 2 :213-19, 1981. [197] D . R o d b a r d a n d A . C h r a m b a c k . E s t i m a t i o n of mo lecu lar mass, free m o b i l i t y , a n d valence us ing p o l y a c r y l a m i d e gel e lectrophoresis . Anal. Biochem., 40(1): 95 -134 , 1971. [198] Y . C o h e n , O . R a m o n , I. J . K o p e l m a n , a n d S. M i z r a h i . C h a r a c t e r i z a t i o n of inhomogeneous p o l y a c r y l a m i d e hydrogels . J. Polymer Sci.: Part B: Polymer Phys., 30(9) :1055-1067, 1992. [199] P . G . R i g h e t t i , B . C . W . B r o s t , a n d R . S. Snyder . O n the l i m i t i n g pore size of h y d r o p h i l i c gels for electrophoresis a n d isoelectr ic focusing. J. Biochem. Biophys. Meth., 4 :347-363, 1981. [200] R . R u c h e l , R . L . Steere, a n d E . F . E r b e . T r a n s m i s s i o n e lectron microscop ic observations of freeze etched p o l y a c r y l a m i d e gels. J. Chromatogr., 166 :563 -575, 1978. [201] R . R u c h e l a n d M . D . Brager . S c a n n i n g e lectron microscope observations of p o l y a c r y l a m i d e gels. Anal. Biochem., 68 :415-428, 1975. [202] G . O d i a n . Principles of Polymerization. J o h n W i l e y a n d Sons Inc. , N e w Y o r k , N Y , 1991. [203] I m P A C T . I m P A C T Dose Survey C T D I R e s u l t s . T e c h n i c a l report , I m P A C T G r o u p , 2004. [204] H . H . B a r r e t t , S . K . G o r d o n , a n d R . S . Hershe l . S t a t i s t i c a l l i m i t a t i o n s i n t r a n s a x i a l tomography . Comp. Biol. Med., 6:307, 1976. [205] M . M . G o o d s i t t a n d R . H . J o h n s o n . P r e c i s i o n i n q u a n t i t a t i v e C T : I m p a c t of x - r a y dose a n d m a t r i x size. Med. Phys., 19:1025-36, 1992. [206] R . C . Gonza lez a n d R . E . W o o d s . Digital Image Processing., P r e n t i c e - H a l l , U p p e r Saddle R i v e r N J , 2001. [207] S. M . S m i t h a n d J . M . B r a d y . S U S A N - a new a p p r o a c h to low level image processing. Int. Journal of Computer Vision, 2 3 ( l ) : 4 5 - 7 8 , 1997. [208] R . T . C h i n a n d C . Y e h . Q u a n t i t a t i v e eva luat i on of some edge-preserving noise-s m o o t h i n g techniques. Computer Vision, Graphics and Image Processing, 23: 6 7 - 9 1 , 1983. [209] M . J . M c D o n n e l l . B o x - f i l t e r i n g techniques. Computer graphics and image processing, 17:65-70, 1981. [210] J . B . B e d n a r a n d T . L . W a t t . A l p h a - t r i m m e d means a n d the i r r e la t i onsh ip to m e d i a n filters. IEEE Trans, on Acoustics, Speech and Signal Processing, 32 (1) :145-153, 1984. [211] J . Lee . D i g i t a l image s m o o t h i n g a n d the s i g m a fi lter . Computer Vision and Image Processing, 24 :255-69, 1983. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0085737/manifest

Comment

Related Items