Digital Volumetric Biopsy Cores Improve Gleason Grading of Prostate Cancer Using Deep Learning

TL;DR

This study introduces a novel 3D volumetric biopsy analysis method using deep learning, significantly improving prostate cancer Gleason grading accuracy over traditional 2D slice-based approaches.

Contribution

It presents a new digital pathology data source called volumetric cores and a deep learning framework that leverages 3D tissue structure for better grading accuracy.

Findings

Achieved a macro-average AUC of 0.958 in GGG classification.

Outperformed 2D baseline models in accuracy and F1 score.

Constructed over 10,000 volumetric cores with a novel alignment framework.

Abstract

Prostate cancer (PCa) was the most frequently diagnosed cancer among American men in 2023. The histological grading of biopsies is essential for diagnosis, and various deep learning-based solutions have been developed to assist with this task. Existing deep learning frameworks are typically applied to individual 2D cross-sections sliced from 3D biopsy tissue specimens. This process impedes the analysis of complex tissue structures such as glands, which can vary depending on the tissue slice examined. We propose a novel digital pathology data source called a "volumetric core," obtained via the extraction and co-alignment of serially sectioned tissue sections using a novel morphology-preserving alignment framework. We trained an attention-based multiple-instance learning (ABMIL) framework on deep features extracted from volumetric patches to automatically classify the Gleason Grade Group (GGG). To handle volumetric patches, we used a modified video transformer with a deep feature extractor pretrained using self-supervised learning. We ran our morphology-preserving alignment framework to construct 10,210 volumetric cores, leaving out 30% for pretraining. The rest of the dataset was used to train ABMIL, which resulted in a 0.958 macro-average AUC, 0.671 F1 score, 0.661 precision, and 0.695 recall averaged across all five GGG significantly outperforming the 2D baselines.