Optimal Fractionation Scheduling for Radiotherapy Treatments with Reinforcement Learning, Tumor Growth Modeling and Outcome Modeling DOI Creative Commons

Mélanie Ghislain,

F. A. Martin,

Manon Dausort

и другие.

Biomedicines, Год журнала: 2025, Номер 13(6), С. 1367 - 1367

Опубликована: Июнь 3, 2025

Objective: Radiotherapy is a primary method for cancer treatment, wherein radiation doses are divided into multiple sessions or fractions to effectively target tumors and minimize damage surrounding tissues. Methods: In this study, we leverage reinforcement learning (RL) enhance treatment planning with the aim of improving adaptability robustness RL agents given inherent inaccuracies in tumor growth models. A 2D simulation model employed, where tabular techniques used determine optimal strategies. We emphasize significance tissue predictions incorporate Lyman NTCP assess outcomes, analyzing complications across three simulated body sites: rectum, head neck lung. Results: For all sites, approach significantly reduces healthy by 10.7%, 49.1% 37.5%, respectively, rectal, lung cancers compared baseline treatment. Conclusions: The RL-based radiotherapy not only achieves eradication but also traditional methods. This study demonstrates potential optimize radiotherapy, offering promising path towards more personalized effective treatments.

Язык: Английский

Optimal Fractionation Scheduling for Radiotherapy Treatments with Reinforcement Learning, Tumor Growth Modeling and Outcome Modeling DOI Creative Commons

Mélanie Ghislain,

F. A. Martin,

Manon Dausort

и другие.

Biomedicines, Год журнала: 2025, Номер 13(6), С. 1367 - 1367

Опубликована: Июнь 3, 2025

Objective: Radiotherapy is a primary method for cancer treatment, wherein radiation doses are divided into multiple sessions or fractions to effectively target tumors and minimize damage surrounding tissues. Methods: In this study, we leverage reinforcement learning (RL) enhance treatment planning with the aim of improving adaptability robustness RL agents given inherent inaccuracies in tumor growth models. A 2D simulation model employed, where tabular techniques used determine optimal strategies. We emphasize significance tissue predictions incorporate Lyman NTCP assess outcomes, analyzing complications across three simulated body sites: rectum, head neck lung. Results: For all sites, approach significantly reduces healthy by 10.7%, 49.1% 37.5%, respectively, rectal, lung cancers compared baseline treatment. Conclusions: The RL-based radiotherapy not only achieves eradication but also traditional methods. This study demonstrates potential optimize radiotherapy, offering promising path towards more personalized effective treatments.

Язык: Английский

Процитировано

0