期刊
FRONTIERS IN NEUROSCIENCE
卷 13, 期 -, 页码 -出版社
FRONTIERS MEDIA SA
DOI: 10.3389/fnins.2019.00097
关键词
cerebrovascular disease; deep learning; medical imaging; segmentation; U-net
资金
- German Federal Ministry of Education and Research through grant Centre for Stroke Research Berlin
- German Federal Ministry of Education and Research through Go-Bio grant for the research group PREDICTioN2020
- German Research Foundation (DFG) via the Open Access Publication Fund of Charite - Universitatsmedizin Berlin
Brain vessel status is a promising biomarker for better prevention and treatment in cerebrovascular disease. However, classic rule-based vessel segmentation algorithms need to be hand-crafted and are insufficiently validated. A specialized deep learning method-the U-net -is a promising alternative. Using labeled data from 66 patients with cerebrovascular disease, the U-net framework was optimized and evaluated with three metrics: Dice coefficient, 95% Hausdorff distance (95HD) and average Hausdorff distance (AVD). The model performance was compared with the traditional segmentation method of graph-cuts. Training and reconstruction was performed using 2D patches. A full and a reduced architecture with less parameters were trained. We performed both quantitative and qualitative analyses. The U-net models yielded high performance for both the full and the reduced architecture: A Dice value of similar to 0.88, a 95HD of similar to 47 voxels and an AVD of similar to 0.4 voxels. The visual analysis revealed excellent performance in large vessels and sufficient performance in small vessels. Pathologies like cortical laminar necrosis and a rete mirabile led to limited segmentation performance in few patients. The U-net outperfomed the traditional graph-cuts method (Dice similar to 0.76, 95HD similar to 59, AVD similar to 1.97). Our work highly encourages the development of clinically applicable segmentation tools based on deep learning. Future works should focus on improved segmentation of small vessels and methodologies to deal with specific pathologies.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据