Scenario Dose Calculation for Robust Optimization in Proton Therapy Treatment Planning
(2021) In Master’s Theses in Mathematical Sciences FMAM05 20211Mathematics (Faculty of Engineering)
- Abstract
- Robust optimization plays a vital role in proton therapy treatment planning. A monoenergetic beam of protons has a finite range and delivers most of its dose at some specific depth in the patient. This allows the design of treatments that mainly damage the tumor and not any surrounding tissue. However, this property also makes proton therapy sensitive to uncertainties in the patient model used for treatment planning. For example, the patient might be slightly misaligned during the treatment and the delivered dose then deviates from the plan. Robust optimization samples such uncertainties into scenarios and simultaneously optimizes the treatment for all scenarios.
This thesis studies an approximate scenario dose calculation method with... (More) - Robust optimization plays a vital role in proton therapy treatment planning. A monoenergetic beam of protons has a finite range and delivers most of its dose at some specific depth in the patient. This allows the design of treatments that mainly damage the tumor and not any surrounding tissue. However, this property also makes proton therapy sensitive to uncertainties in the patient model used for treatment planning. For example, the patient might be slightly misaligned during the treatment and the delivered dose then deviates from the plan. Robust optimization samples such uncertainties into scenarios and simultaneously optimizes the treatment for all scenarios.
This thesis studies an approximate scenario dose calculation method with the potential to improve the computational efficiency of robust optimization in proton therapy treatment planning. The method uses the concept of water-equivalent path lengths to calculate scenario doses as range-corrected deformations of the nominal dose. The method is implemented in a commercial treatment planning system and compared to both a Monte Carlo baseline and an approximate alternative. Compared to the approximate alternative, the proposed method handles a wider class of scenarios which also includes patient setup rotations and anatomical changes.
The results indicate that the proposed method offers no improvements in terms of accuracy. However, when used in the context of robust optimization, no significant degradation of the optimized treatment plan quality is observed in three realistic patient cases. This indicates that the method might be useful when efficiency is the bottleneck. The method is not optimized for performance within the scope of this thesis. However, based on the results and a theoretical analysis, the largest performance gains can be expected for pencil beam scanning plans with many spots per beam. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9059444
- author
- Sundström, Johan LU
- supervisor
- organization
- course
- FMAM05 20211
- year
- 2021
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- cancer, radiation therapy, proton therapy, treatment planning, robust optimization, dose calculation algorithms, scenarios, patient setup, range error, organ motion, pencil beam scanning, computational efficiency, range-corrected dose-mapping
- publication/series
- Master’s Theses in Mathematical Sciences
- report number
- LUTFMA-3449-2021
- ISSN
- 1404-6342
- other publication id
- 2021:E32
- language
- English
- id
- 9059444
- date added to LUP
- 2021-08-20 16:09:48
- date last changed
- 2021-08-20 16:09:48
@misc{9059444, abstract = {{Robust optimization plays a vital role in proton therapy treatment planning. A monoenergetic beam of protons has a finite range and delivers most of its dose at some specific depth in the patient. This allows the design of treatments that mainly damage the tumor and not any surrounding tissue. However, this property also makes proton therapy sensitive to uncertainties in the patient model used for treatment planning. For example, the patient might be slightly misaligned during the treatment and the delivered dose then deviates from the plan. Robust optimization samples such uncertainties into scenarios and simultaneously optimizes the treatment for all scenarios. This thesis studies an approximate scenario dose calculation method with the potential to improve the computational efficiency of robust optimization in proton therapy treatment planning. The method uses the concept of water-equivalent path lengths to calculate scenario doses as range-corrected deformations of the nominal dose. The method is implemented in a commercial treatment planning system and compared to both a Monte Carlo baseline and an approximate alternative. Compared to the approximate alternative, the proposed method handles a wider class of scenarios which also includes patient setup rotations and anatomical changes. The results indicate that the proposed method offers no improvements in terms of accuracy. However, when used in the context of robust optimization, no significant degradation of the optimized treatment plan quality is observed in three realistic patient cases. This indicates that the method might be useful when efficiency is the bottleneck. The method is not optimized for performance within the scope of this thesis. However, based on the results and a theoretical analysis, the largest performance gains can be expected for pencil beam scanning plans with many spots per beam.}}, author = {{Sundström, Johan}}, issn = {{1404-6342}}, language = {{eng}}, note = {{Student Paper}}, series = {{Master’s Theses in Mathematical Sciences}}, title = {{Scenario Dose Calculation for Robust Optimization in Proton Therapy Treatment Planning}}, year = {{2021}}, }