Benchmarking Self-Supervised Learning on Diverse Pathology Datasets

TL;DR

This study conducts the largest-scale benchmarking of self-supervised learning methods on pathology datasets, demonstrating superior performance over traditional pre-training and introducing domain-specific techniques for improved results, including nuclei segmentation.

Contribution

It provides a comprehensive comparison of SSL methods in pathology, proposes domain-specific improvements, and applies SSL successfully to nuclei segmentation tasks.

Findings

Large-scale domain-aligned pre-training outperforms ImageNet pre-training.

Domain-specific techniques enhance SSL performance.

SSL significantly improves nuclei instance segmentation.

Abstract

Computational pathology can lead to saving human lives, but models are annotation hungry and pathology images are notoriously expensive to annotate. Self-supervised learning has shown to be an effective method for utilizing unlabeled data, and its application to pathology could greatly benefit its downstream tasks. Yet, there are no principled studies that compare SSL methods and discuss how to adapt them for pathology. To address this need, we execute the largest-scale study of SSL pre-training on pathology image data, to date. Our study is conducted using 4 representative SSL methods on diverse downstream tasks. We establish that large-scale domain-aligned pre-training in pathology consistently out-performs ImageNet pre-training in standard SSL settings such as linear and fine-tuning evaluations, as well as in low-label regimes. Moreover, we propose a set of domain-specific techniques that we experimentally show leads to a performance boost. Lastly, for the first time, we apply SSL to the challenging task of nuclei instance segmentation and show large and consistent performance improvements under diverse settings.