Cerebrovascular segmentation from TOF-MRA based on multiple-U-net with focal loss function.
Xiaoyu GuoRuoxiu XiaoYuanyuan LuCheng ChenFei YanKangneng ZhouWanzhang HeZhiliang WangPublished in: Comput. Methods Programs Biomed. (2021)
Keyphrases
- loss function
- magnetic resonance angiography
- vessel segmentation
- blood vessels
- support vector
- logistic regression
- pairwise
- phase contrast
- region growing
- learning to rank
- image segmentation
- regularization term
- segmentation method
- convex loss functions
- level set
- markov random field
- similarity measure
- infrared
- total variation
- image restoration
- segmentation algorithm
- carotid artery
- risk minimization
- medical images
- empirical risk
- multiresolution