ESTRO 2025 - Abstract Book

S2457

Physics - Autosegmentation

ESTRO 2025

Conclusion: Overall, the three auto-segmentation models showed similar performance, and all were challenged by both branching organs and organs in areas with lower soft tissue contrast. The auto-segmentation tools exhibited higher consistency compared to inter-observer variations.

Keywords: thorax, comparison, deep-learning

References: [1] Kong F-M (Spring), Ritter T, Quint DJ, Senan S, Gaspar LE, Komaki RU, et al. Consideration of Dose Limits for Organs at Risk of Thoracic Radiotherapy: Atlas for Lung, Proximal Bronchial Tree, Esophagus, Spinal Cord, Ribs, and Brachial Plexus. Int J Radiat Oncol 2011;81:1442–57. https://doi.org/10.1016/j.ijrobp.2010.07.1977. [2] R.T.O.G. Atlases for Organs at Risk (OARs) in Thoracic Radiation Therapy n.d. [3] Yang J, Veeraraghavan H, Armato SG, Farahani K, Kirby JS, Kalpathy ‐ Kramer J, et al. Autosegmentation for thoracic radiation treatment planning: A grand challenge at AAPM 2017. Med Phys 2018;45:4568–81. https://doi.org/10.1002/mp.13141.

2014

Digital Poster Evaluating Inter-Expert Variability and Model Performance in Automated Brain Metastasis Segmentation Anne-Charlotte Roux 1 , Laurie Marchi 1 , Alexis Perrot 1 , Stéphanie Rudzinka 1 , François Lucia 1 , Brieg Dissaux 1 , Thomas Theodoridis 2 , Alexis Bombezin-Domino 3 , Anne Walrafen 3 , Sami Romdhani 2 , Gizem Temiz 3 , Nikos Paragios 4 , Vincent Bourbonne 1,5 1 Radiology Department, University Hospital Brest, Brest, France. 2 AI Engineering, Therapanacea, Paris, France. 3 Clinical Affairs, Therapanacea, Paris, France. 4 CEO, Therapanacea, Paris, France. 5 GETBO, University of Brest, Brest, France