SRA: A Novel Method to Improve Feature Embedding in Self-supervised Learning for Histopathological Images

Hamid Manoochehri, Bodong Zhang, Beatrice Knudsen, Tolga Tasdizen

Keywords: image augmentation, histopathological image, self-supervised learning, contrastive learning

TL;DR: We propose a novel image augmentation method on H&E stained histopathological images that further improves the original self-supervised learning (SSL) models.

Abstract: Self-supervised learning has become a cornerstone in various areas, particularly histopathological image analysis. Image augmentation plays a crucial role in self-supervised learning, as it generates variations in image samples. However, traditional image augmentation techniques often overlook the unique characteristics of histopathological images. In this paper, we propose a new histopathology-specific image augmentation method called stain reconstruction augmentation (SRA). We integrate our SRA into various self-supervised learning models. We demonstrate that our SRA always improves the standard models across various downstream tasks and achieves superior performance to a state-of-the-art foundation model pre-trained on significantly larger histopathology datasets.

Submission Number: 97