Problem
Research questions and friction points this paper is trying to address.
T1-weighted MRI
ischemic stroke lesions
intensity standardization
deep learning architectures
multi-center
Native-Space 3D CarveMix for Multi-Site T1w Stroke Segmentation
🤖 AI Summary
针对多中心T1加权MRI中缺血性卒中病灶分割难题，提出结合MedNeXt-L骨干与3D CarveMix增强方法，动态生成合成病灶，提高分割精度。
📝 Abstract
Segmenting ischemic stroke lesions on T1-weighted (T1w) MRI acquired across different scanners and protocols without intensity standardization is difficult because lesions are subtle and share intensity characteristics with cerebrospinal fluid. Standard deep learning architectures trained across multiple centers plateau around Dice 0.66, with acute lesions ($\le 7$ days post-stroke) performing substantially worse due to severe sample scarcity. We combine a MedNeXt-L ($k=5$) backbone with on-the-fly 3D CarveMix augmentation that pastes real lesion patches into healthy brain regions during training. By generating synthetic lesion placements dynamically within each fold with subject-level split isolation, the model sees more diverse lesion patterns without requiring pre-generated copies on disk. We evaluate on 1,453 native T1w scans from 55 clinical centers in the ISLES 2026 challenge. Our method achieves a mean 5-fold cross-validation Dice of 0.648 at 500 epochs, a +0.018 improvement over the MedNeXt-L backbone at a matched training budget (0.630)
Innovation
Methods, ideas, or system contributions that make the work stand out.
MedNeXt-L
3D CarveMix
on-the-fly augmentation
synthetic lesion placements
cross-center segmentation
🔎 Similar Papers
No similar papers found.
