ESTRO 2025 - Abstract Book

S3274

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

ESTRO 2025

Conclusion: This study demonstrates the feasibility of reliable sCT generation on artifact-affected images, with 3D information enabling improved synthesis. This supports the potential for MR-only radiotherapy in brain and head-and-neck patients, including those with artifacts on MRI. However, further dose studies are needed to confirm the reliability of sCT for radiotherapy. The approach could also be extended to patients with prostheses facing similar signal void challenges in MRI.

Keywords: Dental artifacts, synthetic CT, deep learning

References: [1] Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., ... & Bengio, Y. (2020). Generative adversarial networks. Communications of the ACM , 63 (11), 139-144. [2] Texier, B., Hémon, C., Queffélec, A., Dowling, J., Bessieres, I., Greer, P., ... & Nunes, J. C. (2024). 3D Unsupervised deep learning method for magnetic resonance imaging-to-computed tomography synthesis in prostate radiotherapy. Physics and Imaging in Radiation Oncology , 31 , 100612. [3] Hémon, C., Texier, B., Chourak, H., Simon, A., Bessières, I., de Crevoisier, R., ... & Nunes, J. C. (2024). Indirect deformable image registration using synthetic image generated by unsupervised deep learning. Image and Vision Computing , 105143.

2648

Proffered Paper Fast intrafraction drift corrections for prostate SBRT on a 1.5T MR-linac Hans C de Boer, Eline N de Groot-van Breugel, Louk M Snoeren, Lieke T Meijers, Bas W Raaymakers, Jochem R van der Voort van Zijp Radiation Oncology, UMC Utrecht, Utrecht, Netherlands Purpose/Objective: Intrafraction motion is a limiting factor in margin reduction in prostate cancer treatment. Recently, comprehensive motion management (CMM) has become available on the 1.5 T MR-linac (Elekta AB), that allows for both gating as well as baseline shift corrections. Here, we report on its application and margin reduction achieved. Material/Methods: Because prostate motion is non-periodic and erratic, we first investigated the appropriate actions levels to be applied in CMM. The cine-MR stream consists of interleaved T2-TSE sagittal and coronal 2D images, acquired at 0.5 Hz. Using these images, the CMM software tracks the 3D position of the prostate relative to the daily 3D MR scan used for treatment planning. CMM can gate the beam if the momentaneous excursion exceeds a certain value. In addition, if a persistent baseline shift is observed, a correction for this shift can be applied to the remaining plan segments. We determined the efficacy of action levels for gating and shift correction based on simulations using CMM imaging in 35 prostate cancer patients treated with 5 x 7.25 Gy (SBRT). The overall goal was to reduce planning