A new deep learning approach for the retinal hard exudates detection based on superpixel multi-feature extraction and patch-based CNN.
Chenxi HuangYongshuo ZongYimin DingXin LuoKathy ClawsonYonghong PengPublished in: Neurocomputing (2021)
Keyphrases
- deep learning
- retinal images
- diabetic retinopathy
- optic disk
- fundus images
- feature extraction
- blood vessels
- optic disc
- vessel segmentation
- unsupervised learning
- automatic detection
- machine learning
- optical coherence tomography
- unsupervised feature learning
- weakly supervised
- region of interest
- feature vectors
- pattern recognition
- image processing
- feature selection
- early detection
- principal component analysis
- feature space
- face recognition
- superpixels
- dimensionality reduction
- medical experts