Neural Stain-Style Transfer Learning using GAN for Histopathological Images
This work addresses stain-style variability in histopathological image analysis for medical diagnosis, representing an incremental improvement by adapting existing GAN methods to this domain-specific issue.
The authors tackled the problem of tumor classification performance varying with stain-styles in histopathological images by proposing a stain-style transfer model based on conditional GANs, which preserved tumor classifier accuracy on transferred images as validated on the CAMELYON16 dataset.
Performance of data-driven network for tumor classification varies with stain-style of histopathological images. This article proposes the stain-style transfer (SST) model based on conditional generative adversarial networks (GANs) which is to learn not only the certain color distribution but also the corresponding histopathological pattern. Our model considers feature-preserving loss in addition to well-known GAN loss. Consequently our model does not only transfers initial stain-styles to the desired one but also prevent the degradation of tumor classifier on transferred images. The model is examined using the CAMELYON16 dataset.