Whole Slide Image Multi-Classification of Cervical Epithelial Lesions Based on Unsupervised Pre-training.
Minfan ZhaoMin LingZhaohui WangJun ShiHongyu KanHong AnWenting HanJoseph BartlettWenqi LuPublished in: EMBC (2022)
Keyphrases
- image classification
- supervised learning
- image data
- image analysis
- digital mammograms
- unsupervised learning
- single image
- multiscale
- image content
- training phase
- input image
- training set
- training process
- image features
- microscopic images
- supervised classification
- training samples
- preprocessing stage
- decision trees
- machine learning
- image retrieval
- segmentation method
- textural features
- hyperspectral images
- multi layer perceptron
- set of training images
- image segmentation
- skin lesion
- markov random field
- training data
- region of interest
- classification accuracy
- image regions
- support vector machine
- high resolution
- test images
- microscopy images
- discriminative classifiers
- restricted boltzmann machine
- similarity measure
- segmentation algorithm
- computer aided detection
- supervised training
- semi supervised
- image representation