Decoding fibrosis: Transcriptomic and clinical insights via AI-derived collagen deposition phenotypes in MASLD

Histological assessment is foundational to multi-omics studies of liver disease, yet conventional fibrosis staging lacks resolution, and quantitative metrics like collagen proportionate area (CPA) fail to capture tissue architecture. While recent artificial intelligence (AI)-driven approaches offer improved precision, they are proprietary and not accessible to academic research.

Here, we present a novel, interpretable AI-based framework for characterizing liver fibrosis from picrosirius red (PSR)-stained slides. By identifying distinct data-driven collagen deposition phenotypes (CDPs) that capture distinct morphologies, our method substantially improves the sensitivity and biological specificity of downstream transcriptomic and proteomic analyses compared with CPA and traditional fibrosis scores. Pathway analysis reveals that CDPs 4 and 5 are associated with active extracellular matrix remodeling, while phenotype correlates highlight links to liver functional status. Importantly, selected CDPs demonstrated prognostic associations in the discovery cohort, with attenuation of discrimination in the external validation cohort.

We've developed a novel digital pathology framework for liver fibrosis quantification. All models and tools are made freely available to support transparent and reproducible multi-omics pathology research.