Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Intrafractional prostate movement studied by electronic portal imaging in cine mode

Månsson Haskå, Torbjörn (2007)
Medical Physics Programme
Abstract (Swedish)
Syftet med denna studie var att utvärdera säkerhetsmarginalerna som läggs kring ett målorgan (tex prostata) vid extern strålterapi. Resultaten kan senare ligga till grund för reducering av säkerhetsmarginalerna, detta för att skydda strål-känslig frisk vävnad från onödig bestrålning. En delstudie var också att försöka finna ett mönster hos prostatans rörelse vid de inledande behandlingstillfällena som kunde vara till hjälp för att beskriva rörelserna under den resterande behandlingen. Detta skulle kunna vara användbart vid dosplanläggningen för utformning av individuella dosplaner för varje patient.

Material och metoder:
20 patienter med prostatacancer deltog i studien och alla hade tre guldstifts-markörer inopererade i prostatan.... (More)
Syftet med denna studie var att utvärdera säkerhetsmarginalerna som läggs kring ett målorgan (tex prostata) vid extern strålterapi. Resultaten kan senare ligga till grund för reducering av säkerhetsmarginalerna, detta för att skydda strål-känslig frisk vävnad från onödig bestrålning. En delstudie var också att försöka finna ett mönster hos prostatans rörelse vid de inledande behandlingstillfällena som kunde vara till hjälp för att beskriva rörelserna under den resterande behandlingen. Detta skulle kunna vara användbart vid dosplanläggningen för utformning av individuella dosplaner för varje patient.

Material och metoder:
20 patienter med prostatacancer deltog i studien och alla hade tre guldstifts-markörer inopererade i prostatan. Markörerna används som surrogat för prostatans position i de digitala filmsekvenser som spelats in i realtid med hjälp av be-handlingsstrålen och en digital bildplatta. Digitala filmsekvenser spelades in under 10 av de 39 behandlingstillfällen som normalt ges vid behandling av prostatacancer. Det utfördes två olika analyser: en omfattande analys av 5 patienter och en förkortad ana-lys av 15 patienter. I den omfattande analysen följdes prostatakörtelns rörelse genom alla inspelade filmsekvenser men i den förkortade studerades endast den första bilden i varje filmsekvens. Ett behandlingstillfälle består av en fraktion som i sin tur består av 5 fält (5 filmsekvenser) och som bestrålas från fem olika gantryvinklar. Till bildanalys och positionsberäkning användes bildbehandlingsprogrammet Image J.

Resultat:
Det visade sig att det endast gick att räkna ut exakta resultat i longitudinell riktning på grund av de valda gantryvinklarna. Så beräkningarna av anterior-posterior riktninge blir endast en grov uppskattning som inte går att använda till att reducera säkerhetsmarginalen. Den omfattande analysen visade att prostata rörde sig väldigt lite under själva bestrålningen av ett fält, men totalt under en hel fraktion kunde prostatan röra sig flera millimeter relativt startpositionen i fält 1. Den största rörelsen som obser-verades i longitudinell riktning var ca 7,5 mm. I 95 % av alla fraktionerna registrera-des en rörelse < 4,1 mm och den största förflyttningen i förhållande till startpositionen i fält 1 observerades ofta i fält 4. Resultaten visade också att prostatan rör sig diagonalt i huvud-mag riktningen. Det fanns inget mönster i ett tidigt skede av behandlingen som kunde beskriva prostatakörtelns rörelse under resterande behandlingstillfällen.

Slutsats:
Av de totalt 196 fraktioner som studerades hade 95 % en rörelse som var < 4,1 mm i longitudinell riktning och medelförflyttningen i varje fält ökade relativt start-positionen fram till fält 4 för att sedan plana ut. Detta innebär att tiden inte påverkar storleken av förflyttningen. Det visade sig samtidigt att de största förflyttningarna oftast skedde i fält 4 vilket ytterligare stärker antagandet om att tiden inte påverkar hur stor förflyttningen kommer att bli. I de fall där det skedde en extremt stor förflyttning av prostatan observerades samtidigt stora mängder gas i tarmen. Det visade sig att prostatans vandring sker slumpmässigt och omöjlig att förutspå. (Less)
Abstract
"Introduction: This thesis is a study of the margins that are necessary in prostate cancer radio therapy treatment. The aim was to analyse the intrafraction motion of the prostate gland during treatment, by the use of portal imaging, in order to optimize the margin around the prostate gland and spare healthy tissue. One of the purposes was to find a common pattern in prostate movement in the first fractions of the treatment that could describe the movement during the following fractions. The intention was to create individual margins for each patient for the rest of the treatment. Further analyses of the portal images were made to visualize the directions and size of the prostate gland motion in order to reduce the planning target volume... (More)
"Introduction: This thesis is a study of the margins that are necessary in prostate cancer radio therapy treatment. The aim was to analyse the intrafraction motion of the prostate gland during treatment, by the use of portal imaging, in order to optimize the margin around the prostate gland and spare healthy tissue. One of the purposes was to find a common pattern in prostate movement in the first fractions of the treatment that could describe the movement during the following fractions. The intention was to create individual margins for each patient for the rest of the treatment. Further analyses of the portal images were made to visualize the directions and size of the prostate gland motion in order to reduce the planning target volume (PTV) already from the first treatment.

Materials and Methods: Twenty patients with diagnosed prostate cancer had three gold seed markers inserted in the prostate gland. A five field forward planned step-and-shoot technique was used for the treatment. The gantry angles used were 180°, 103°, 36°, 257° and 324°. For all patients CINE picture sequences were recorded with the electronic portal imaging devices (EPID) during fractions 2, 3, 4, 9, 14, 19, 24, 29, 34, and 39. On five of the patients a full analysis was performed which consisted of an analysis of all CINE pictures recorded. It was done in order to visualize the magnitude of the motion of the prostate in the three directions lateral (LT), cranial-caudal (CC) and posterior-anterior (PA). On fifteen patients a reduced analysis was performed which contained an analysis of the first CINE picture for each field. This was done in order to visualize the magnitude of the motion of the prostate gland motion in the CC direction. The image analysis was performed in two computer software programs: Image J and Microsoft Excel. A accelerator arm sag measurement was performed on the three different accelerators that were used for treatment. The results of these measurements were used to correct the CC analysis and to estimate a correction factor that could be used in the PA analysis.

Results: There was no common pattern found that could describe the intra¬frac¬tion motion of the prostate gland during the treatment. A specificity and sensitivity analysis showed that it is possible to predict only two thirds of all cases. The results from the full analysis showed that there were only small intrafield displacements (< 1 mm) relative to the intrafraction displacements which were much larger, in some cases several mm. The results from the two analyses showed that there was a correlation in the motion of the prostate gland in the cranial-anterior direction and in the caudal-posterior direction. For one patient the squared correlation factor was 0.75. The size of the motion never exceeded 7 mm in the CC direction, 95 % of all fractions had a maximum motion of < 4.1 mm in the CC direction and 98 % of all field 1’s were < 1.0 mm in the LT direction. No exact results for the anterior-posterior direction were reached.

Conclusions: There were 196 fractions analysed and in 95% of these a deviation of < 4.1 mm in CC direction was observed. An increased mean displacement relative the start position was shown for fields 1, 2, 3 and 4 and then it looks like a steady state was reached. In most cases the maximum displacement in each fraction occurred in field 4. These results imply that the motion of the prostate gland probably reaches a steady state with time. Intrafractional and intrafield motion of the prostate gland is not possible to predict, it can occur in any fraction during treatment and often in combination with gas in the rectum.
" (Less)
Please use this url to cite or link to this publication:
author
Månsson Haskå, Torbjörn
supervisor
organization
year
type
H2 - Master's Degree (Two Years)
subject
keywords
Strålterapi
language
English
id
2157025
date added to LUP
2011-09-14 12:05:31
date last changed
2013-09-05 10:23:02
@misc{2157025,
  abstract     = {{"Introduction: This thesis is a study of the margins that are necessary in prostate cancer radio therapy treatment. The aim was to analyse the intrafraction motion of the prostate gland during treatment, by the use of portal imaging, in order to optimize the margin around the prostate gland and spare healthy tissue. One of the purposes was to find a common pattern in prostate movement in the first fractions of the treatment that could describe the movement during the following fractions. The intention was to create individual margins for each patient for the rest of the treatment. Further analyses of the portal images were made to visualize the directions and size of the prostate gland motion in order to reduce the planning target volume (PTV) already from the first treatment.

Materials and Methods: Twenty patients with diagnosed prostate cancer had three gold seed markers inserted in the prostate gland. A five field forward planned step-and-shoot technique was used for the treatment. The gantry angles used were 180°, 103°, 36°, 257° and 324°. For all patients CINE picture sequences were recorded with the electronic portal imaging devices (EPID) during fractions 2, 3, 4, 9, 14, 19, 24, 29, 34, and 39. On five of the patients a full analysis was performed which consisted of an analysis of all CINE pictures recorded. It was done in order to visualize the magnitude of the motion of the prostate in the three directions lateral (LT), cranial-caudal (CC) and posterior-anterior (PA). On fifteen patients a reduced analysis was performed which contained an analysis of the first CINE picture for each field. This was done in order to visualize the magnitude of the motion of the prostate gland motion in the CC direction. The image analysis was performed in two computer software programs: Image J and Microsoft Excel. A accelerator arm sag measurement was performed on the three different accelerators that were used for treatment. The results of these measurements were used to correct the CC analysis and to estimate a correction factor that could be used in the PA analysis.

Results: There was no common pattern found that could describe the intra¬frac¬tion motion of the prostate gland during the treatment. A specificity and sensitivity analysis showed that it is possible to predict only two thirds of all cases. The results from the full analysis showed that there were only small intrafield displacements (< 1 mm) relative to the intrafraction displacements which were much larger, in some cases several mm. The results from the two analyses showed that there was a correlation in the motion of the prostate gland in the cranial-anterior direction and in the caudal-posterior direction. For one patient the squared correlation factor was 0.75. The size of the motion never exceeded 7 mm in the CC direction, 95 % of all fractions had a maximum motion of < 4.1 mm in the CC direction and 98 % of all field 1’s were < 1.0 mm in the LT direction. No exact results for the anterior-posterior direction were reached.

Conclusions: There were 196 fractions analysed and in 95% of these a deviation of < 4.1 mm in CC direction was observed. An increased mean displacement relative the start position was shown for fields 1, 2, 3 and 4 and then it looks like a steady state was reached. In most cases the maximum displacement in each fraction occurred in field 4. These results imply that the motion of the prostate gland probably reaches a steady state with time. Intrafractional and intrafield motion of the prostate gland is not possible to predict, it can occur in any fraction during treatment and often in combination with gas in the rectum.
"}},
  author       = {{Månsson Haskå, Torbjörn}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Intrafractional prostate movement studied by electronic portal imaging in cine mode}},
  year         = {{2007}},
}