GANs vs. Diffusion Models for virtual staining with the HER2match dataset

2025-06-23

AI Summary

To address the scarcity of publicly available paired data and the lack of consensus on optimal generative models for H&E-to-HER2 virtual staining, this paper introduces HER2match—the first high-quality, publicly accessible paired dataset for this task—and pioneers the application of Brownian Bridge Diffusion Models (BBDMs) to H&E-to-HER2 translation. Through systematic benchmarking of state-of-the-art GANs and diffusion models—including BBDMs—our experiments reveal that while GANs generally achieve superior overall performance, BBDMs attain generation quality comparable to top-tier GANs while preserving high sampling efficiency. Models trained on HER2match significantly improve staining fidelity and histological structural consistency. HER2match thus establishes the first authoritative benchmark dataset for breast cancer pathology virtual staining, and BBDM-based translation provides an efficient, clinically viable generative paradigm—advancing interpretable AI-assisted diagnosis in digital pathology.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Humans and AI: Other Foundations of Human Computation & AI

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Virtual staining is a promising technique that uses deep generative models to recreate histological stains, providing a faster and more cost-effective alternative to traditional tissue chemical staining. Specifically for H&E-HER2 staining transfer, despite a rising trend in publications, the lack of sufficient public datasets has hindered progress in the topic. Additionally, it is currently unclear which model frameworks perform best for this particular task. In this paper, we introduce the HER2match dataset, the first publicly available dataset with the same breast cancer tissue sections stained with both H&E and HER2. Furthermore, we compare the performance of several Generative Adversarial Networks (GANs) and Diffusion Models (DMs), and implement a novel Brownian Bridge Diffusion Model for H&E-HER2 translation. Our findings indicate that, overall, GANs perform better than DMs, with only the BBDM achieving comparable results. Furthermore, we emphasize the importance of data alignment, as all models trained on HER2match produced vastly improved visuals compared to the widely used consecutive-slide BCI dataset. This research provides a new high-quality dataset ([available upon publication acceptance]), improving both model training and evaluation. In addition, our comparison of frameworks offers valuable guidance for researchers working on the topic.

Problem

• Lack of public datasets for H&E-HER2 staining transfer
• Unclear best model frameworks for H&E-HER2 translation
• Need for improved data alignment in virtual staining models

Innovation

• Introduces HER2match dataset for H&E-HER2 staining
• Compares GANs and Diffusion Models performance
• Proposes novel Brownian Bridge Diffusion Model