Generating Seamless Virtual Immunohistochemical Whole Slide Images with Content and Color Consistency

Immunohistochemical (IHC) stains play a vital role in a pathologist's analysis of medical images, providing crucial diagnostic information for various diseases. However, obtaining IHC staining is not always readily available and often costly. Virtual staining from hematoxylin and eosin (H&E)-stained whole slide images (WSIs) produces useful IHC stains without the expensive physical staining process. However, current virtual WSI generation methods based on tile-wise processing often suffer from inconsistencies in content, texture’ and color at tile boundaries. These inconsistencies lead to artifacts that compromise image quality and potentially hinder the adoption of virtual WSI and accurate clinical assessment. To address this limitation, this paper proposes a novel consistent whole slide image (WSI) synthesis network, CC-WSI-Net, which generates seamless and realistic synthetic WSIs from hematoxylin and eosin (H&E)-stained images. Our method addresses the limitations of current virtual WSI generation methods by integrating a content and color consistency supervisor, ensuring consistency across tiles and accurate detection of clinically relevant features. CC-WSI-Net is validated via image quality analysis, detection assessments, and a survey conducted with pathologists. Notably, in 67% of cases, pathologists could not distinguish between traditionally stained slides and the synthetic slides from CC-WSI-Net.