ESTRO 2025 - Abstract Book

S3283

Physics - Intra-fraction motion management and real-time adaptive radiotherapy

ESTRO 2025

Conclusion: The bowel tracking framework was used to detect significant displacements from the reference position in two ways. First, the largest voxel-wise difference in position, showed that in all cases half of the voxels experience a difference of at least 5 mm in craniocaudal direction and on average almost 9 mm. Second, the drift over time of the mean positions within the bowel bag, showing that the main drift is in craniocaudal direction, where in most cases the shift is in cranial direction after 15 minutes. Given the found range of motion, we foresee that long-term bowel displacements may have an impact on the delivered dose and therefore, such information can be used in applications such as dose accumulation or the determination of PRV margins.

Keywords: Intrafraction bowel motion, MR-guided radiotherapy

References: 1. Berthold Horn and Brian Schunck. Determining optical flow. Artificial intelligence , 17:185-203, 08 1981. 2. Danique L.J. Barten, Janna J. Laan, Koen J. Nelissen, Jorrit Visser, Henrike Westerveld, Arjan Bel, Catharina S. de Jonge, Jaap Stoker, and Zdenko van Kesteren. A 3d cine-mri acquisition technique and image analysis framework to quantify bowel motion demonstrated in gynecological cancer patients. Medical Physics , 48(6):3109-3119, 2021.

2773

Digital Poster Interplay-robust optimization for treating irregularly breathing patients with pencil beam scanning Ivar Bengtsson 1,2 , Anders Forsgren 1 , Albin Fredriksson 2 , Ye Zhang 3 1 Mathematics, KTH Royal Institute of Technology, Stockholm, Sweden. 2 Research, RaySearch Laboratories, Stockholm, Sweden. 3 Center for proton therapy, Paul Scherrer Institut, Villigen, Switzerland Purpose/Objective: Robust optimization (RO) has been successfully implemented to mitigate patient positioning and density uncertainties for proton pencil beam scanning (PBS). Although RO can be extended to motion uncertainty (4DRO) by considering multiple images (e.g. from a 4DCT) as scenarios, its clinical application to motion management is typically accompanied by other motion management techniques such as gating and rescanning [1]. As such techniques increase treatment times or patient discomfort, it is interesting whether motion management can be mitigated efficiently already in treatment plan optimization.