Histopathology DatasetGAN: Synthesizing Large-Resolution Histopathology Datasets

TL;DR

This paper introduces HDGAN, a scalable semi-supervised framework for generating and segmenting high-resolution histopathology images, addressing the challenge of applying self-supervised learning to large medical datasets.

Contribution

HDGAN extends DatasetGAN with adaptations like a new generator backbone and memory-efficient features, enabling effective high-resolution medical image synthesis and segmentation.

Findings

HDGAN achieves strong performance on high-resolution histopathology datasets.

The framework reduces memory consumption, making it suitable for large medical images.

Demonstrates potential for broader application of deep learning in medical imaging.

Abstract

Self-supervised learning (SSL) methods are enabling an increasing number of deep learning models to be trained on image datasets in domains where labels are difficult to obtain. These methods, however, struggle to scale to the high resolution of medical imaging datasets, where they are critical for achieving good generalization on label-scarce medical image datasets. In this work, we propose the Histopathology DatasetGAN (HDGAN) framework, an extension of the DatasetGAN semi-supervised framework for image generation and segmentation that scales well to large-resolution histopathology images. We make several adaptations from the original framework, including updating the generative backbone, selectively extracting latent features from the generator, and switching to memory-mapped arrays. These changes reduce the memory consumption of the framework, improving its applicability to medical imaging domains. We evaluate HDGAN on a thrombotic microangiopathy high-resolution tile dataset, demonstrating strong performance on the high-resolution image-annotation generation task. We hope that this work enables more application of deep learning models to medical datasets, in addition to encouraging more exploration of self-supervised frameworks within the medical imaging domain.