Automated polyp segmentation in colonoscopy images via deep network with lesion-aware feature selection and refinement.
Guanghui YueWanwan HanSiying LiTianwei ZhouJun LvTianfu WangPublished in: Biomed. Signal Process. Control. (2022)
Keyphrases
- segmentation algorithm
- pigmented skin lesions
- feature selection
- segmentation method
- automated segmentation
- image analysis
- skin lesion
- test images
- fully automatic
- segmentation accuracy
- edge detection
- image data
- pixel wise
- region of interest
- ground truth
- image regions
- adaptive thresholding
- input image
- image features
- medical images
- image retrieval
- segmented images
- pet images
- automatically segmented
- segmentation errors
- accurate segmentation
- grey level
- image registration
- mri images
- microscopic images
- pixel level
- multiple objects
- image segments
- level set
- dermoscopy images
- positron emission tomography
- image segmentation
- automated analysis
- object segmentation
- region growing
- multiscale
- cell nuclei
- ct images
- eye tracking
- mutual information
- feature set
- image sequences
- automated image analysis