Feature Selection and Reduction for Analysis of Histopathology Images

Applications ranging from computer assisted diagnosis (CAD) to image categorization and retrieval are very much popular in the automatic analysis of microscopic images. The research community is increasingly measuring certain key properties in images, such as tissue architecture, color, texture, and morphology. Extracting feature and selecting a vital feature set is an important task in CAD. The removal of undesirable elements and interferences from surface histopathology images is accomplished using the important technique of feature extraction. Picking a feature vector needs to be done carefully if the goal is to successfully classify the biopsy images. However, an array of features containing a lot of redundant characteristics are being employed in many investigations of the histopathology image classification. This work presents a brief summary of the key algorithms and approaches for feature extraction and selection in histopathological images. The major goal of this research is to bring together a comprehensive overview of the computational techniques used to quantify visual aspects in histology images.