An empirical survey on Uterine cancer Tissue classification using Histopathological Images

Abstract

Histopathological images (HIs) are the gold standard for determining cancer diagnoses in specific cases. Even for expert pathologists, analyzing such images takes time and resources, and it is difficult, resulting in inter-observer and intra-observer discrepancies. Using computer-aided diagnostic (CAD) technologies is one technique to speed up such an analysis. Machine learning methods for histopathology image analysis, including shallow and deep learning methods, are discussed in this work. We also go over preprocessing, segmentation and feature extraction, three of the most common tasks in histopathological image analysis. The classification job has increased interest, whereas other tasks such as segmentation and feature extraction have shown a fall in interest for tissue diagnosis. This empirical study depicts the pros and cons of the various preprocessing, segmentation, classification and feature extraction methods in the histopathological images for the effective implementation purposes.