Stain Isolation-based Guidance for Improved Stain Translation
This work addresses stain translation issues in histopathology imaging, which is incremental as it builds on existing CycleGAN methods with a new loss function.
The paper tackles the problem of non-structure-preserving errors in unsupervised stain translation of histopathology images by introducing a guidance scheme based on stain isolation consistency, showing improved translation between IHC and mIF domains in qualitative and quantitative experiments.
Unsupervised and unpaired domain translation using generative adversarial neural networks, and more precisely CycleGAN, is state of the art for the stain translation of histopathology images. It often, however, suffers from the presence of cycle-consistent but non structure-preserving errors. We propose an alternative approach to the set of methods which, relying on segmentation consistency, enable the preservation of pathology structures. Focusing on immunohistochemistry (IHC) and multiplexed immunofluorescence (mIF), we introduce a simple yet effective guidance scheme as a loss function that leverages the consistency of stain translation with stain isolation. Qualitative and quantitative experiments show the ability of the proposed approach to improve translation between the two domains.