ESTRO 2025 - Abstract Book

S2442

Physics - Autosegmentation

ESTRO 2025

Keywords: cervical cancer, auto-contouring evaluation

References: 1.

Mackay K, Bernstein D, Glocker B, Kamnitsas K, Taylor A. A Review of the Metrics Used to Assess Auto Contouring Systems in Radiotherapy. Clin Oncol (R Coll Radiol) . Jan 31 2023;doi:10.1016/j.clon.2023.01.016 2. Bernstein D, Taylor A, Nill S, Oelfke U. New target volume delineation and PTV strategies to further personalise radiotherapy. Phys Med Biol . Feb 25 2021;66(5):055024. doi:10.1088/1361-6560/abe029 3. Isensee F, Jaeger PF, Kohl SAA, Petersen J, Maier-Hein KH. nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nat Methods . Feb 2021;18(2):203-211. doi:10.1038/s41592-020 01008-z

1713

Proffered Paper A foundation model for head and neck organs-at-risk segmentation in CT scans: toward universal segmentation from clinically partial annotations Wenjun Liao 1 , Xiangde Luo 2 , He Li 2 , Shichuan Zhang 1 1 Radiation Oncology, Sichuan Cancer Hospital & Institute, Chengdu, China. 2 School of Mechanical and Electrical Engineering, University of Electronic Science and Technology of China, Chengdu, China Purpose/Objective: Foundation models (FMs) have garnered significant attention due to their exceptional performance. However, the effectiveness of the powerful deep learning-based FMs often hinges on large-scale, high-quality datasets. In the task of head and neck cancer (HNC) organs-at-risk (OARs) segmentation, obtaining such datasets poses a substantial challenge. To address this issue, we collected a large-scale partially labeled dataset comprising HNC patients and developed a FM capable of accurately delineating 43 OARs for HNC radiotherapy.

Material/Methods: