Automated Gleason grading of prostate cancer tissue microarrays via deep learning

TLDR

The Gleason grading system remains the most powerful prognostic predictor for prostate cancer, yet its application requires highly trained pathologists, is tedious, and suffers from limited inter‑pathologist reproducibility, especially for intermediate score 7. Automated annotation procedures offer a viable solution to address these limitations. We developed a deep‑learning model trained on detailed Gleason annotations from 641 patients and evaluated on an independent cohort of 245 patients annotated by two pathologists. The model achieved inter‑annotator agreement of 0.75 and 0.71 with the pathologists, comparable to their 0.71 agreement, and produced expert‑level patient stratification into prognostically distinct groups, demonstrating promise for objective, reproducible grading of heterogeneous Gleason patterns.

Abstract

Abstract The Gleason grading system remains the most powerful prognostic predictor for patients with prostate cancer since the 1960s. Its application requires highly-trained pathologists, is tedious and yet suffers from limited inter-pathologist reproducibility, especially for the intermediate Gleason score 7. Automated annotation procedures constitute a viable solution to remedy these limitations. In this study, we present a deep learning approach for automated Gleason grading of prostate cancer tissue microarrays with Hematoxylin and Eosin (H&E) staining. Our system was trained using detailed Gleason annotations on a discovery cohort of 641 patients and was then evaluated on an independent test cohort of 245 patients annotated by two pathologists. On the test cohort, the inter-annotator agreements between the model and each pathologist, quantified via Cohen’s quadratic kappa statistic, were 0.75 and 0.71 respectively, comparable with the inter-pathologist agreement (kappa = 0.71). Furthermore, the model’s Gleason score assignments achieved pathology expert-level stratification of patients into prognostically distinct groups, on the basis of disease-specific survival data available for the test cohort. Overall, our study shows promising results regarding the applicability of deep learning-based solutions towards more objective and reproducible prostate cancer grading, especially for cases with heterogeneous Gleason patterns.