Improving Cnns Classification With Pathologist-based Expertise: The Renal Cell Carcinoma Case Study

This study presents a novel strategy for automatic RCC subclassification, referred to as ExpertDeepTree (ExpertDT). ExpertDT combines supervised deep learning models (CNNs) and pathologist’s expertise to boost the system’s performance in very cumbersome classification cases. Results from experiments on a dataset of 201 WSIs from 91 patients show that the ExpertDT approach has superior ability in the RCC subtyping task.