ESTRO 2025 - Abstract Book

S3773

Physics - Radiomics, functional and biological imaging and outcome prediction

ESTRO 2025

Conclusion: Random Forest demonstrates statistical superiority with robust metrics for predicting post-neoadjuvant recurrence. The proposed methodology establishes a reproducible framework for validating predictive models in radiotherapy oncology, enabling risk stratification for personalized therapy.

Keywords: Machine learning, RAPIDO, Recurrence prediction

References: Nilsson, P., & al., et al. (2021). Total Neoadjuvant Therapy with Short-Course Radiation in Locally Advanced Rectal Cancer (RAPIDO): A Randomized Phase III Trial. The Lancet Oncology , 22(7), 947-957. https://doi.org/10.1016/S1470 2045(21)00353-4