Advanced

Optical surface scanning of breast cancer patients in radiotherapy – an investigation of inter- and intrafractional motion effects

Mannerberg, Annika (2017) MSFT01 20171
Medical Physics Programme
Abstract
Purpose: The overall purpose of this master thesis was to investigate if the patient positioning could be improved for breast cancer patients with nodal involvement, using the optical scanning system Catalyst during patient setup. Patient motion during treatment and the effect it has on the dose distribution was also investigated.

Materials and Methods: Eleven patients positioned according to the Catalyst and 10 patients positioned according to the conventional laser and skin markings based setup (LBS) were enrolled in this study. To evaluate if positioning was improved with the surface based setup (SBS), the setup deviations arising from matching daily verification images to the reference images were acquired for both positioning... (More)
Purpose: The overall purpose of this master thesis was to investigate if the patient positioning could be improved for breast cancer patients with nodal involvement, using the optical scanning system Catalyst during patient setup. Patient motion during treatment and the effect it has on the dose distribution was also investigated.

Materials and Methods: Eleven patients positioned according to the Catalyst and 10 patients positioned according to the conventional laser and skin markings based setup (LBS) were enrolled in this study. To evaluate if positioning was improved with the surface based setup (SBS), the setup deviations arising from matching daily verification images to the reference images were acquired for both positioning methods. The setup deviations were then compared between the different methods by studying the distribution of deviations. The systematic and random setup deviations and PTV margin were also determined for both groups. In total, 127 fractions of patients positioned with LBS and 93 fractions of patients positioned with SBS were analysed. To investigate what impact the offline correction strategy used at Skåne University Hospital (SUS) has on the positioning, the data was analysed both with and without the offline corrections applied.
The dosimetric effect of patient motion was evaluated for one patient by analysing recorded motion data during beam delivery by the Catalyst system in terms of isocenter shift. All isocenter shifts that occurred when the beam was on were extracted and added to the original patient position to get the total isocenter shift, including both inter- and intrafractional motion. The mean isocenter shift for each fraction was determined and the isocenter position was shifted in the original treatment plan for each fraction in the treatment planning system (TPS). The dose was then re-calculated for each fraction and summed to one plan that was compared to the original dose distribution.

Results: The obtained results indicate that SBS is a better method of positioning patients than LBS. For LBS, with offline corrections included, the amount of fractions where setup deviation exceeded the clinical setup deviation threshold of 4 mm was 22% / 21% / 28% in vrt/lng/lat. The corresponding values for SBS were 7.5% / 6.5% / 20% in vrt/lng/lat. The systematic and random setup error was smaller with Catalyst setup in all directions and the PTV margin could be reduced in the lng direction for SBS, compared to LBS. The setup deviations were also evaluated for LBS and SBS, both with and without the offline corrections applied. The results showed that although it is important to use the correction strategy for LBS, SBS did not depend on the correction strategy and the positioning was not improved as much for this method. No large impact on the DVH could be seen for the PTV of the summed treatment plan where the isocenter was shifted to a mean position for each fraction. D98% of the original plan was 46.6 Gy. D98% for the plan where the isocenter was shifted was 46.5 Gy. The absorbed dose to the heart and left lung was lower than in the original plan and the absorbed dose to the spinal cord was slightly increased. The difference between the plans was not clinically relevant.

Conclusions: The amount of setup deviations exceeding 4 mm was decreased in all directions for SBS compared to LBS. Also, both the systematic and random setup error was decreased for SBS. The correction strategy improved the positioning for LBS, while the SBS did not depend as much on the corrections. The DVH of one patient with considerable motion during treatment indicated that the dose distribution was not affected to any large extent, when inter- and intrafractional motion was considered. However, to be able to draw any definite conclusions about the dosimetric effect due to patient motion more patients have to be evaluated. (Less)
Popular Abstract (Swedish)
Förbättrad positionering av bröstcancerpatienter som genomgår strålbehandling

Cirka 30 000 kvinnor får diagnosen cancer varje år i Sverige [1]. Av dessa utgör bröstcancer den vanligaste diagnosen med över 9000 fall per år [2]. Cancern kan vara begränsad till endast bröstet, men kan också ha spridit sig till de intilliggande lymfkörtlarna. Behandling av bröstcancer med körtelengagemang inleds vanligtvis med kirurgi och därefter är det vanligt att patienten även får strålbehandling. Syftet med detta är att, med hjälp av strålningen, eliminera de tumörceller som fortfarande kan finnas kvar efter operationen och på så sätt minska risken för att en ny tumör börjar växa. På Skånes Universitetssjukhus i Lund delas strålbehandlingen för... (More)
Förbättrad positionering av bröstcancerpatienter som genomgår strålbehandling

Cirka 30 000 kvinnor får diagnosen cancer varje år i Sverige [1]. Av dessa utgör bröstcancer den vanligaste diagnosen med över 9000 fall per år [2]. Cancern kan vara begränsad till endast bröstet, men kan också ha spridit sig till de intilliggande lymfkörtlarna. Behandling av bröstcancer med körtelengagemang inleds vanligtvis med kirurgi och därefter är det vanligt att patienten även får strålbehandling. Syftet med detta är att, med hjälp av strålningen, eliminera de tumörceller som fortfarande kan finnas kvar efter operationen och på så sätt minska risken för att en ny tumör börjar växa. På Skånes Universitetssjukhus i Lund delas strålbehandlingen för bröstcancerpatienter med körtelengagemang upp på 25 fraktioner. Patienten får en lika stor stråldos vid varje behandlingstillfälle.

Innan strålbehandlingen påbörjas genomgår patienten datortomografi., som genom röntgenteknik genererar snittbilder av patienten. Vid detta tillfälle positioneras patienten i en fixation som ska användas under alla kommande behandlingstillfällen. Fixationen säkerställer att patienten ligger still och likadant under hela behandlingen. Utifrån datortomografin görs sedan en stråldosplan, i vilken man bestämmer hur strålningen ska ges och med vilka maskininställningar. Vid strålbehandling är det viktigt att endast tumörområdet bestrålas och att den friska vävnaden skonas i så stor utsträckning som möjligt. En viktig del för att uppnå detta utgörs av att patienten ligger korrekt positionerad varje dag vid behandlingen. Det konventionella sättet att positionera patienten inför behandling är att använda sig av ett laserbaserat koordinatsystem som finns inne i behandlingsrummet samt markeringar som ritas på patientens hud vid datortomografin. När laser och hudmarkeringar sammanfaller antar man att patienten är korrekt positionerad. För att verifiera patientens position tas röntgenbilder på patienten innan behandlingen startas. Detta görs de tre första fraktionerna och därefter en gång i veckan. Röntgenbilderna matchas mot referensbilder från CT-skanningen och när dessa bilder överlappar vet man att patienten kommer få sin behandling till rätt område. De dagar som verifikationsbilder inte tas, positionerar man patienten endast utifrån laser och hudmarkeringar. Trots att patienten är fixerad finns det utrymme för rörelse. Patientrörelser kan påverka tumörområdets position, utan att detta kan upptäckas med laser och hudmarkeringar. För bröstcancerpatienter påverkas bestrålningsvolymens position exempelvis av armens position, som är placerad ovanför huvudet. Eftersom inga hudmarkeringar ritas på patientens armar finns det risk för att en sådan felpositionering inte upptäcks och patienten får inte den stråldos till den volym som var planerat.

För att förbättra patientpositioneringen under strålbehandling har optiskt ytskanning utvecklats, vilket bygger på att synligt ljus skannar av en vald volym av patientens yta medan hen ligger på britsen. Med hjälp av det reflekterade ljuset från patienten kan systemet beräkna patientens position i rummet och därmed upptäcka en eventuell felpositionering av patienten. Ljuset som patienten skannas med är vanligt, synligt ljus och ger därför ingen extra stråldos. Eftersom det är en hel volym av patienten som skannas kan exempelvis felpositionerade armar hos bröstcancerpatienter upptäckas och justeras.

I detta examensarbete undersöktes om patientpositioneringen av bröstcancerpatienter med körtelengagemang kunde förbättras genom att positionera patienterna med ett optiskt ytskanningssystem. Studien genomfördes genom att patienter positionerades med detta optiska ytskanningssystem, istället för med laser och hudmarkeringar. För alla fraktioner där röntgenbilder togs noterades avvikelsen mellan dagens bild och referensbilden. Dessa siffror jämfördes därefter med motsvarande siffror från patienter som hade lagts upp med hjälp av laser och hudmarkeringar för att undersöka om optisk ytskanning bidrog till mindre avvikelser i positioneringen.

Studiens resultat visar på en förbättring av patientpositioneringen då optisk ytskanning används. Detta innebär att patienten kan positioneras med en god noggrannhet även de dagar då verifikationsbilder inte tas, utan att erhålla någon extra stråldos från röntgenbildtagning. Med hjälp av bättre daglig patientpositionering får patienten även en bättre behandling eftersom rätt volym blir bestrålad. (Less)
Please use this url to cite or link to this publication:
author
Mannerberg, Annika
supervisor
organization
course
MSFT01 20171
year
type
H2 - Master's Degree (Two Years)
subject
language
English
id
8926415
date added to LUP
2017-09-27 20:41:42
date last changed
2017-09-28 09:11:29
@misc{8926415,
  abstract     = {Purpose: The overall purpose of this master thesis was to investigate if the patient positioning could be improved for breast cancer patients with nodal involvement, using the optical scanning system Catalyst during patient setup. Patient motion during treatment and the effect it has on the dose distribution was also investigated.

Materials and Methods: Eleven patients positioned according to the Catalyst and 10 patients positioned according to the conventional laser and skin markings based setup (LBS) were enrolled in this study. To evaluate if positioning was improved with the surface based setup (SBS), the setup deviations arising from matching daily verification images to the reference images were acquired for both positioning methods. The setup deviations were then compared between the different methods by studying the distribution of deviations. The systematic and random setup deviations and PTV margin were also determined for both groups. In total, 127 fractions of patients positioned with LBS and 93 fractions of patients positioned with SBS were analysed. To investigate what impact the offline correction strategy used at Skåne University Hospital (SUS) has on the positioning, the data was analysed both with and without the offline corrections applied.
The dosimetric effect of patient motion was evaluated for one patient by analysing recorded motion data during beam delivery by the Catalyst system in terms of isocenter shift. All isocenter shifts that occurred when the beam was on were extracted and added to the original patient position to get the total isocenter shift, including both inter- and intrafractional motion. The mean isocenter shift for each fraction was determined and the isocenter position was shifted in the original treatment plan for each fraction in the treatment planning system (TPS). The dose was then re-calculated for each fraction and summed to one plan that was compared to the original dose distribution.

Results: The obtained results indicate that SBS is a better method of positioning patients than LBS. For LBS, with offline corrections included, the amount of fractions where setup deviation exceeded the clinical setup deviation threshold of 4 mm was 22% / 21% / 28% in vrt/lng/lat. The corresponding values for SBS were 7.5% / 6.5% / 20% in vrt/lng/lat. The systematic and random setup error was smaller with Catalyst setup in all directions and the PTV margin could be reduced in the lng direction for SBS, compared to LBS. The setup deviations were also evaluated for LBS and SBS, both with and without the offline corrections applied. The results showed that although it is important to use the correction strategy for LBS, SBS did not depend on the correction strategy and the positioning was not improved as much for this method. No large impact on the DVH could be seen for the PTV of the summed treatment plan where the isocenter was shifted to a mean position for each fraction. D98% of the original plan was 46.6 Gy. D98% for the plan where the isocenter was shifted was 46.5 Gy. The absorbed dose to the heart and left lung was lower than in the original plan and the absorbed dose to the spinal cord was slightly increased. The difference between the plans was not clinically relevant.

Conclusions: The amount of setup deviations exceeding 4 mm was decreased in all directions for SBS compared to LBS. Also, both the systematic and random setup error was decreased for SBS. The correction strategy improved the positioning for LBS, while the SBS did not depend as much on the corrections. The DVH of one patient with considerable motion during treatment indicated that the dose distribution was not affected to any large extent, when inter- and intrafractional motion was considered. However, to be able to draw any definite conclusions about the dosimetric effect due to patient motion more patients have to be evaluated.},
  author       = {Mannerberg, Annika},
  language     = {eng},
  note         = {Student Paper},
  title        = {Optical surface scanning of breast cancer patients in radiotherapy – an investigation of inter- and intrafractional motion effects},
  year         = {2017},
}