Multi-Organ Cancer Classification and Survival Analysis

Accurate and robust cell nuclei classification is the cornerstone for a wider range of tasks in digital and Computational Pathology. However, most machine learning systems require extensive labeling from expert pathologists for each individual problem at hand, with no or limited abilities for knowledge transfer between datasets and organ sites. In this paper we implement and evaluate a variety of deep neural network models and model ensembles for nuclei classification in renal cell cancer (RCC) and prostate cancer (PCa). We propose a convolutional neural network system based on residual learning which significantly improves over the state-of-the-art in cell nuclei classification. However, the main thrust of our work is to demonstrate for the first time the models ability to transfer the learned concepts from one organ type to another, with robust performance. Finally, we show that the combination of tissue types during training increases not only classification accuracy but also overall survival analysis. The best model, trained on combined data of RCC and PCa, exhibits optimal performance on PCa classification and better survival group stratification than an expert pathologist ($p=0.006$). All code, image data and expert labels are made publicly available to serve as benchmark for the community for future research into computational pathology.