Cascaded multitask U-Net using topological loss for vessel segmentation and centerline extraction.
Pierre RougéNicolas PassatOdyssée MerveillePublished in: CoRR (2023)
Keyphrases
- vessel segmentation
- multi task
- blood vessels
- multitask learning
- retinal images
- multi task learning
- low contrast
- learning tasks
- region growing
- vessel enhancement
- multi class
- multiple tasks
- gaussian processes
- feature selection
- transfer learning
- learning problems
- information extraction
- region of interest
- information gain
- topology preserving
- support vector machine
- image segmentation
- machine learning algorithms
- edge detection
- supervised learning
- reinforcement learning