Two-level training of a 3D U-Net for accurate segmentation of the intra-cochlear anatomy in head CTs with limited ground truth training data.
Dongqing ZhangRueben A. BanalagayJianing WangYiyuan ZhaoJack H. NobleBenoit M. DawantPublished in: Medical Imaging: Image Processing (2019)
Keyphrases
- accurate segmentation
- ground truth
- training data
- training set
- automatic segmentation
- mr images
- training examples
- medical images
- supervised learning
- low contrast
- active shape model
- training samples
- high quality
- prior knowledge
- learning algorithm
- decision trees
- three dimensional
- magnetic resonance
- pairwise
- maximum likelihood
- segmentation algorithm
- semi automatic
- image processing