PanNuke: An Open Pan-Cancer Histology Dataset for Nuclei Instance Segmentation and Classification.
Jevgenij GamperNavid Alemi KoohbananiKsenija BenetAli KhuramNasir M. RajpootPublished in: ECDP (2019)
Keyphrases
- benchmark datasets
- segmentation method
- pattern recognition
- microscopic images
- feature set
- feature space
- cancer diagnosis
- breast cancer
- segmentation algorithm
- gene expression datasets
- classification accuracy
- training set
- support vector
- decision trees
- cancer datasets
- image classification
- gene expression profiles
- image segmentation
- pixel classification
- microscopy images
- cancer classification
- cell nuclei
- prostate cancer
- text classification
- gene expression data
- feature vectors
- feature extraction
- cancer cells
- confocal images
- colon cancer
- lymph nodes
- machine learning
- features extraction
- fully unsupervised
- lung cancer
- training dataset
- object segmentation
- supervised learning
- object detection
- image analysis
- feature selection