Detection of multiple retinal diseases in ultra-widefield fundus images using deep learning: data-driven identification of relevant regions.
Justin EngelmannAlice D. McTrustyIan J. C. MacCormickEmma PeadAmos J. StorkeyMiguel O. BernabeuPublished in: CoRR (2022)
Keyphrases
- fundus images
- deep learning
- data driven
- diabetic retinopathy
- blood vessels
- optic disc
- optical coherence tomography
- retinal images
- automatic detection
- vascular tree
- unsupervised learning
- retinal blood vessels
- feature selection
- early detection
- region of interest
- co occurrence
- medical experts
- domain specific
- active learning