UE 1.1 Medical Image Analysis

Medical image acquisition & features

Methodology design in medical image analysis

Medical image segmentation

• binary vs semantic vs instance segmentation
• active contours (snake)
• evaluation metrics for medical image segmentation
• more "old school" (ie unsupervised techniques)

Deep learning in medical image segmentation

• from classification networks to segmentation networks
• pioneering networks: FCN, UNet
• various architectures
• loss functions (cross-entropy, dice)
• mitigate the need for labeled data
  • data augmentation in the training set: with geometric transformation or generative models
  • use weakly labeled or unlabeled data with weakly supervised learning, semi-supervised learning

Image registration

Characterization of images

• Characterization methods (Statistical attributes, Co-occurrence matrix, Mutlifractal analysis, Filtering, Representation of shape
• Feature extraction with auto-encoder

Multimodal medical image fusion

• Information Fusion (Fuzzy sets, Belief functions, Probability theory)
• Deep learning based fusion