Mitotic v/s non-mitotic HEp-2 cells classification for CAD based auto-immune disorder detection
Krati Gupta, Arnav V. Bhavsar, Anil Kumar Sao · 2018
We propose a HEp-2 cell classification framework for mitotic v/s non-mitotic cells for detection of autoimmune disorders. The important aspects of such classification framework is to use Convolutional Neural Network (CNN) as learned feature extractor, as to incorporate the prominent and distinctive shape and texture differences within the cells. Considering the data sample skew issue between the mitotic and non-mitotic cells and giving insights based on the performance of traditional Support Vector Machine (SVM) and CNN based classifiers over learned feature representations, we demonstrate a high classification performance with proposed framework. The approach has been evaluated on a publicly available dataset of individual mitotic cells.