English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Renal pathology, as the gold standard of kidney disease diagnosis, requiresdoctors to analyze a series of tissue slices stained by H&E staining andspecial staining like Masson, PASM, and PAS, respectively. These specialstaining methods are costly, time-consuming, and hard to standardize for wideuse especially in primary hospitals. Advances of supervised learning methodshave enabled the virtually conversion of H&E images into special stainingimages, but achieving pixel-to-pixel alignment for training remainschallenging. In contrast, unsupervised learning methods regarding differentstains as different style transfer domains can utilize unpaired data, but theyignore the spatial inter-domain correlations and thus decrease thetrustworthiness of structural details for diagnosis. In this paper, we proposea novel virtual staining framework AGMDT to translate images into other domainsby avoiding pixel-level alignment and meanwhile utilizing the correlationsamong adjacent tissue slices. We first build a high-quality multi-domain renalhistological dataset where each specimen case comprises a series of slicesstained in various ways. Based on it, the proposed framework AGMDT discoverspatch-level aligned pairs across the serial slices of multi-domains throughglomerulus detection and bipartite graph matching, and utilizes suchcorrelations to supervise the end-to-end model for multi-domain stainingtransformation. Experimental results show that the proposed AGMDT achieves agood balance between the precise pixel-level alignment and unpaired domaintransfer by exploiting correlations across multi-domain serial pathologicalslices, and outperforms the state-of-the-art methods in both quantitativemeasure and morphological details.

AGMDT: Virtual Staining of Renal Histology Images with Adjacency-Guided Multi-Domain Transfer