MS-SST: Single Image Reconstruction-Based Stain-Style Transfer for Multi-Domain Hematoxylin & Eosin Stained Pathology Images

In digital pathology, pathological tissue images that are obtained using scanners analyzed and diseases diagnosed. One crucial aspect of this process is the staining slides. However, differences appear in color even when same protocol owing to various factors such as different facilities, hospitals, scanning equipment. Many stain style normalization studies have been conducted solve problem. study, we propose a model named multi-domain single image reconstruction-based stain-style transfer. The proposed trained learning framework, which can efficiently reduce complexity training time compared with associated GAN objective. We randomly extracted stained patches from CAMELYON17 Mitos-Atypia-14 datasets demonstrated an effective translation. Our study reveals it possible perform translation among multiple domains per domain. Furthermore, experimentally temperature was natural code publicly available at: https://github.com/jwkweon/MS-SST .