ESTRO 2025 - Abstract Book

S3125

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2025

Conclusion: This study demonstrated the potential of DDPMs incorporated with a CBCT-guided sampling approach to generate high-fidelity sCT images. Our methods significantly improved image quality metrics compared to conventional DDPMs, enabling sCT generation suitable for adaptive radiotherapy workflows. This approach offers a promising solution for clinics with limited datasets, paving the way for more effective and accessible CBCT-based adaptive radiotherapy.

Keywords: synthetic CT, diffusion model, machine learning

References: 1. Ho, J., Jain, A., & Abbeel, P. (2020). Denoising diffusion probabilistic models. Advances in neural information processing systems, 33, 6840-6851. 2. Meng, C., He, Y., Song, Y., Song, J., Wu, J., Zhu, J. Y., & Ermon, S. (2021). Sdedit: Guided image synthesis and editing with stochastic differential equations. arXiv preprint arXiv:2108.01073. 3. Peng, J., Qiu, R. L., Wynne, J. F., Chang, C. W., Pan, S., Wang, T., ... & Yang, X. (2024). CBCT ‐ Based synthetic CT image generation using conditional denoising diffusion probabilistic model. Medical physics, 51(3), 1847-1859.

2123

Digital Poster Assessing the Impact of Hounsfield Unit Variation of HyperSight CBCT on CT-CBCT Deformable Image Registration Yu-chi Hu Medical Physics, Memorial Sloan Kettering Cancer Center, New York, USA