ESTRO 2025 - Abstract Book

S2468

Physics - Autosegmentation

ESTRO 2025

The developed models demonstrated robust performance across all evaluated segmentation tasks. Our approach was ranked first in the PED, second in the MenRT, third in the HNTS-MRG, and leading results in the GLI challenges, underscoring the effectiveness and generalizability of the proposed methodology.

Figure 1. Illustrative examples of segmented tumors for each challenge.

Conclusion: The proposed models outperform current state-of-the-art methods, enabling accurate segmentation of brain and head&neck tumors in MRIs. Next steps include deploying these models in a research PACS for clinical evaluation.

Keywords: glioma, meningioma, head and neck

References: [1] M. C. de Verdier et al. , “The 2024 Brain Tumor Segmentation (BraTS) Challenge: Glioma Segmentation on Post treatment MRI”, May 2024, https://arxiv.org/abs/2405.18368v1 [2] A. F. Kazerooni et al. , “The Brain Tumor Segmentation in Pediatrics (BraTS-PEDs) Challenge: Focus on Pediatrics (CBTN-CONNECT-DIPGR-ASNR-MICCAI BraTS-PEDs)”, Apr. 2024, https://arxiv.org/abs/2404.15009v4 [3] D. LaBella et al. , “Brain Tumor Segmentation (BraTS) Challenge 2024: Meningioma Radiotherapy Planning Automated Segmentation”, May 2024, https://arxiv.org/abs/2405.18383v1 [4] “Head and Neck Tumor Segmentation for MR-Guided Applications (HNTS-MRG) 2024”, https://hntsmrg24.grand challenge.org/

2321

Proffered Paper How to spot, and stop, automation bias in autocontouring Paul J Doolan 1 , Afxentia Karousiou 2 , Djamal Boukerroui 3 , Constantinos Zamboglou 4,5 , Konstantinos Ferentinos 4 , Mark Gooding 3,6 1 Medical Physics, German Oncology Center, Limassol, Cyprus. 2 Medical Physics, University of Surrey, London, United Kingdom. 3 InPictura, Ltd, Abingdon, United Kingdom. 4 Radiation Oncology, German Oncology Center, Limassol, Cyprus. 5 Radiation Oncology, University of Freiberg, Freiberg, Germany. 6 Division of Cancer Sciences, The University of Manchester, Manchester, United Kingdom