ESTRO 2025 - Abstract Book

S3546

Physics - Optimisation, algorithms and applications for ion beam treatment planning

ESTRO 2025

Conclusion: This study demonstrates that Bayesian statistics is a powerful tool for parameter optimization in cell survival models. Even with a small set of data (69), it was possible to robustly adjust parameters without critical dependence on the chosen prior. This procedure improves the prediction of the theoretical model and, in principle, is applicable to various mechanistic models.

Keywords: Survival Fraction, Bayes, PIDE

References: [1] Stewart, R. D. (2001). Radiation research, 156(4), 365-378. doi:10.1667/0033 7587(2001)156[0365:TLKMOD]2.0.CO;2 [2] Wang, W., et al. (2018). Scientific reports, 8(1), 16202. doi:10.1038/s41598-018-34159-3

[3] McMahon, et al. (2016). Scientific reports, 6(1), 33290. doi:10.1038/srep33290 [4] Carlson, D. J., et al. (2008). Radiation research, 169(4), 447-459. doi:10.1667/RR1046.1 [5] Friedrich, T., et al. (2021). Journal of Radiation Research, 62(4), 645-655. doi:10.1093/jrr/rrab034 [6] Semenenko, V. A., et al. (2006). Physics in Medicine & Biology, 51(7), 1693. doi:10.1088/0031-9155/51/7/004 [7] Abril-Pla, et al. (2023). PeerJ Computer Science, 9, e1516 doi:10.7717/peerj-cs.1516

3798

Proffered Paper Fast deliverable proton therapy for lung cancer using 3D-printed range modulators: evaluating beam time efficiency and motion mitigation Joshua Beyers 1 , Kenneth Poels 2 , Willemijn Goossens 2 , Maarten Lambrecht 2,1 , Patrick Berkovic 2 , Antoine Delor 3 , Sima Qamhiyeh 2 , Rasmus Nilsson 4 , Erik Traneus 4 , Tom Depuydt 2,1 1 Laboratorium voor experimentele radiotherapie, KU Leuven, Leuven, Belgium. 2 Radiation Oncology department, UZ Leuven, Leuven, Belgium. 3 Radiotherapie oncologique, Clinique Universitaires Saint-Luc, Brussels, Belgium. 4 Particle Therapy, Raysearch Laboraties AB, Stockholm, Sweden Purpose/Objective: Hedgehog 3D-printed range modulators (3DRM) have the potential to make very fast delivery times possible for PBS PT by avoiding energy switching. Conformal target dose could in this way be realized with one single energy layer. Even without FLASH ultra-high dose rates, this approach can be used for ultra-fast delivery, reducing the impact of tumor motion in the dose delivery. The potential use of 3DRM is demonstrated in a lung cancer planning study assessing plan quality, delivery times and motion interplay effects. Material/Methods: IMPT plans were made for respectively a lung phantom , locally-advanced (LA) (#3) and SBRT lung cancer patients (#8) using single-field (SF) and multi-field (MF) beam arrangements. These plans were converted to 3DRM plans using a modified Raystation FLASH version (v11B, RaysearchLabs) (Fig 1a-1b). Unlike for FLASH where cyclotron E max usually is selected, here the single layer energy was adjusted to the target depth. Similar to the FLASH solution a collimating aperture was used with a margin to the target volume of 1 cm. Delivery times were calculated for the beam with the largest spot count, hereby assuming scanning dwell-times and energy switching of a ProteusOne (IBA). The conformity index (CI) and gradient-index GI (V80%/V95%) were calculated for IMPT plans and for 3DRM plans both aperture-collimated and uncollimated. Using in-house scripting, interplay effects were simulated for SF IMPT and MF-IMPT on a dynamic phantom (2cm p2p-motion) and compared with feasibility of free-breathing gating with 3DRM delivery.