# Deep learning-based bias transfer for overcoming laboratory differences of microscopic images

arXiv: https://arxiv.org/abs/2105.11765 · HTML: https://mlsanity.com/paper/2105.11765/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 9 of its 24 references that are in the map.
- First posted on arXiv 25 May 2021.

## Builds on (most cited first, 9 of 9)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Image-to-Image Translation with Conditional Adversarial Networks](https://mlsanity.com/paper/1611.07004/) (2016)
- [StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation](https://mlsanity.com/paper/1711.09020/) (2017)
- [Distribution Matching Losses Can Hallucinate Features in Medical Image Translation](https://mlsanity.com/paper/1805.08841/) (2018)
- [Cycle-Dehaze: Enhanced CycleGAN for Single Image Dehazing](https://mlsanity.com/paper/1805.05308/) (2018)
- [StainGAN: Stain Style Transfer for Digital Histological Images](https://mlsanity.com/paper/1804.01601/) (2018)
- [Learning Fixed Points in Generative Adversarial Networks: From Image-to-Image Translation to Disease Detection and Localization](https://mlsanity.com/paper/1908.06965/) (2019)
- [Unsupervised Adversarial Correction of Rigid MR Motion Artifacts](https://mlsanity.com/paper/1910.05597/) (2019)
- [Noise as Domain Shift: Denoising Medical Images by Unpaired Image Translation](https://mlsanity.com/paper/1910.02702/) (2019)

## Abstract (arXiv)

The automated analysis of medical images is currently limited by technical and biological noise and bias. The same source tissue can be represented by vastly different images if the image acquisition or processing protocols vary. For an image analysis pipeline, it is crucial to compensate such biases to avoid misinterpretations. Here, we evaluate, compare, and improve existing generative model architectures to overcome domain shifts for immunofluorescence (IF) and Hematoxylin and Eosin (H&E) stained microscopy images. To determine the performance of the generative models, the original and transformed images were segmented or classified by deep neural networks that were trained only on images of the target bias. In the scope of our analysis, U-Net cycleGANs trained with an additional identity and an MS-SSIM-based loss and Fixed-Point GANs trained with an additional structure loss led to the best results for the IF and H&E stained samples, respectively. Adapting the bias of the samples significantly improved the pixel-level segmentation for human kidney glomeruli and podocytes and improved the classification accuracy for human prostate biopsies by up to 14%.