Pathologist-like explainable AI for interpretable Gleason grading in prostate cancer

TL;DR

This paper introduces an explainable AI system for Gleason grading in prostate cancer, using a new annotated dataset and a U-Net architecture to provide interpretable predictions aligned with pathologists' terminology.

Contribution

It presents a novel, annotated dataset and an inherently explainable AI model that outperforms traditional methods, capturing uncertainty and aligning with clinical reasoning.

Findings

AI achieves a Dice score of 0.713 on Gleason pattern segmentation.

The model maintains high performance with high interobserver variability.

The dataset and approach promote further research in subjective medical segmentation tasks.

Abstract

The aggressiveness of prostate cancer, the most common cancer in men worldwide, is primarily assessed based on histopathological data using the Gleason scoring system. While artificial intelligence (AI) has shown promise in accurately predicting Gleason scores, these predictions often lack inherent explainability, potentially leading to distrust in human-machine interactions. To address this issue, we introduce a novel dataset of 1,015 tissue microarray core images, annotated by an international group of 54 pathologists. The annotations provide detailed localized pattern descriptions for Gleason grading in line with international guidelines. Utilizing this dataset, we develop an inherently explainable AI system based on a U-Net architecture that provides predictions leveraging pathologists' terminology. This approach circumvents post-hoc explainability methods while maintaining or exceeding the performance of methods trained directly for Gleason pattern segmentation (Dice score: 0.713 $\pm$ 0.003 trained on explanations vs. 0.691 $\pm$ 0.010 trained on Gleason patterns). By employing soft labels during training, we capture the intrinsic uncertainty in the data, yielding strong results in Gleason pattern segmentation even in the context of high interobserver variability. With the release of this dataset, we aim to encourage further research into segmentation in medical tasks with high levels of subjectivity and to advance the understanding of pathologists' reasoning processes.