Towards Reliable Pediatric Brain Tumor Segmentation: Task-Specific nnU-Net Enhancements
This addresses the critical need for reliable pediatric brain tumor segmentation to aid diagnosis and treatment, though it is incremental as it builds on existing nnU-Net methods.
The paper tackled accurate segmentation of pediatric brain tumors in MRI by developing an enhanced nnU-Net framework, achieving first place on the BraTS 2025 Task-6 validation leaderboard with lesion-wise Dice scores up to 0.967.
Accurate segmentation of pediatric brain tumors in multi-parametric magnetic resonance imaging (mpMRI) is critical for diagnosis, treatment planning, and monitoring, yet faces unique challenges due to limited data, high anatomical variability, and heterogeneous imaging across institutions. In this work, we present an advanced nnU-Net framework tailored for BraTS 2025 Task-6 (PED), the largest public dataset of pre-treatment pediatric high-grade gliomas. Our contributions include: (1) a widened residual encoder with squeeze-and-excitation (SE) attention; (2) 3D depthwise separable convolutions; (3) a specificity-driven regularization term; and (4) small-scale Gaussian weight initialization. We further refine predictions with two postprocessing steps. Our models achieved first place on the Task-6 validation leaderboard, attaining lesion-wise Dice scores of 0.759 (CC), 0.967 (ED), 0.826 (ET), 0.910 (NET), 0.928 (TC) and 0.928 (WT).