# Region-guided CycleGANs for Stain Transfer in Whole Slide Images

arXiv: https://arxiv.org/abs/2208.12847 · HTML: https://mlsanity.com/paper/2208.12847/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 22 references that are in the map.
- First posted on arXiv 26 Aug 2022.

## Builds on (most cited first, 17 of 17)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Mask R-CNN](https://mlsanity.com/paper/1703.06870/) (2017)
- [Image-to-Image Translation with Conditional Adversarial Networks](https://mlsanity.com/paper/1611.07004/) (2016)
- [Perceptual Losses for Real-Time Style Transfer and Super-Resolution](https://mlsanity.com/paper/1603.08155/) (2016)
- [Least Squares Generative Adversarial Networks](https://mlsanity.com/paper/1611.04076/) (2016)
- [Data Efficient and Weakly Supervised Computational Pathology on Whole Slide Images](https://mlsanity.com/paper/2004.09666/) (2020)
- [Coupled Generative Adversarial Networks](https://mlsanity.com/paper/1606.07536/) (2016)
- [Object-driven Text-to-Image Synthesis via Adversarial Training](https://mlsanity.com/paper/1902.10740/) (2019)
- [PathologyGAN: Learning deep representations of cancer tissue](https://mlsanity.com/paper/1907.02644/) (2019)
- [Self-Supervised Representation Learning using Visual Field Expansion on Digital Pathology](https://mlsanity.com/paper/2109.03299/) (2021)
- [GAN-based Virtual Re-Staining: A Promising Solution for Whole Slide Image Analysis](https://mlsanity.com/paper/1901.04059/) (2019)
- [Generative Adversarial Networks in Digital Pathology: A Survey on Trends and Future Potential](https://mlsanity.com/paper/2004.14936/) (2020)
- [Pedestrian-Synthesis-GAN: Generating Pedestrian Data in Real Scene and Beyond](https://mlsanity.com/paper/1804.02047/) (2018)
- [Computational Histological Staining and Destaining of Prostate Core Biopsy RGB Images with Generative Adversarial Neural Networks](https://mlsanity.com/paper/1811.02642/) (2018)
- [Realistic Image Generation using Region-phrase Attention](https://mlsanity.com/paper/1902.05395/) (2019)
- [A Generative Adversarial Model for Right Ventricle Segmentation](https://mlsanity.com/paper/1810.03969/) (2018)

## Built on by (most cited first, 8 of 8)

- [Multistain Pretraining for Slide Representation Learning in Pathology](https://mlsanity.com/paper/2408.02859/) (2024)
- [PixCell: A generative foundation model for digital histopathology images](https://mlsanity.com/paper/2506.05127/) (2025)
- [A comparative evaluation of image-to-image translation methods for stain transfer in histopathology](https://mlsanity.com/paper/2303.17009/) (2023)
- [Topology-Guided Multi-Class Cell Context Generation for Digital Pathology](https://mlsanity.com/paper/2304.02255/) (2023)
- [A novel framework employing deep multi-attention channels network for the autonomous detection of metastasizing cells through fluorescence microscopy](https://mlsanity.com/paper/2309.00911/) (2023)
- [From Pixels to Pathology: Restoration Diffusion for Diagnostic-Consistent Virtual IHC](https://mlsanity.com/paper/2508.02528/) (2025)
- [Artifact Removal in Histopathology Images](https://mlsanity.com/paper/2211.16161/) (2022)
- [Hallucination Detection in Virtually-Stained Histology: A Latent Space Baseline](https://mlsanity.com/paper/2411.15060/) (2024)

## Abstract (arXiv)

In whole slide imaging, commonly used staining techniques based on hematoxylin and eosin (H&E) and immunohistochemistry (IHC) stains accentuate different aspects of the tissue landscape. In the case of detecting metastases, IHC provides a distinct readout that is readily interpretable by pathologists. IHC, however, is a more expensive approach and not available at all medical centers. Virtually generating IHC images from H&E using deep neural networks thus becomes an attractive alternative. Deep generative models such as CycleGANs learn a semantically-consistent mapping between two image domains, while emulating the textural properties of each domain. They are therefore a suitable choice for stain transfer applications. However, they remain fully unsupervised, and possess no mechanism for enforcing biological consistency in stain transfer. In this paper, we propose an extension to CycleGANs in the form of a region of interest discriminator. This allows the CycleGAN to learn from unpaired datasets where, in addition, there is a partial annotation of objects for which one wishes to enforce consistency. We present a use case on whole slide images, where an IHC stain provides an experimentally generated signal for metastatic cells. We demonstrate the superiority of our approach over prior art in stain transfer on histopathology tiles over two datasets. Our code and model are available at https://github.com/jcboyd/miccai2022-roigan.