Generalizable multi-task, multi-domain deep segmentation of sparse pediatric imaging datasets via multi-scale contrastive regularization and multi-joint anatomical priors.
Arnaud BoutillonPierre-Henri ConzeChristelle PonsValérie BurdinBhushan BorotikarPublished in: CoRR (2022)
Keyphrases
- multi task
- multi domain
- multiscale
- group lasso
- sparse learning
- elastic net
- multi task learning
- medical images
- sparsity inducing
- structured sparsity
- medical imaging
- image processing
- image segmentation
- cross domain
- prior information
- learning tasks
- transfer learning
- multiple tasks
- learning problems
- metric learning
- multi class
- sparse representation
- feature selection
- domain specific
- prior knowledge
- gaussian processes
- data sets
- regularization parameter
- natural images
- magnetic resonance images
- machine learning algorithms
- high dimensional
- decision trees
- bayesian framework
- machine learning