Journal article
Accuracy of Foundation Artificial Intelligence Models for Hepatic Macrovesicular Steatosis Quantification in Frozen Sections
Archives of pathology & laboratory medicine (1976), v 150(5), pp 374-380
15 Dec 2025
PMID: 41397435
Abstract
Context - Accurate intraoperative assessment of macrovesicular steatosis in donor liver biopsies is critical for transplant decisions but is often limited by interobserver variability and freezing artifacts that can obscure histologic details. Artificial intelligence (AI) offers a potential solution for standardized and reproducible evaluation.
Objective - To evaluate the diagnostic performance of 2 self-supervised learning (SSL)-based foundation models, Prov-GigaPath and UNI, for classifying macrovesicular steatosis on frozen liver biopsy sections, compared with assessments by surgical pathologists.
Design - This retrospective study included 131 frozen liver biopsy specimens from 68 donors collected between November 2022 and September 2024. Slides were digitized into whole slide images, tiled into patches, and used to extract embeddings with Prov-GigaPath and UNI; slide-level classifiers were then trained and tested. Intraoperative diagnoses by on-call surgical pathologists were compared with ground truth determined from independent reviews of permanent sections by 2 liver pathologists. Accuracy was evaluated for both a 5-category classification and a clinically significant binary threshold (<30% versus >= 30%).
Results - For the binary classification, Prov-GigaPath achieved 96.4% accuracy, UNI 85.7%, and surgical pathologists, 89.3% (P = .37). For the 5-category classification, accuracies were lower: Prov-GigaPath, 57.1%; UNI, 50.0%; and pathologists, 64.2% (P = .47). Misclassification occurred mainly in intermediate categories (5% to <30% steatosis).
Conclusions - SSL-based foundation models performed comparably to surgical pathologists at the clinically relevant threshold of less than 30% versus 30% or greater. These findings support the potential role of AI in standardizing intraoperative evaluation of donor liver biopsies; however, the small sample size limits generalizability and requires validation in larger, balanced cohorts. (Arch Pathol Lab Med. 2026;150:374-380; doi: 10.5858/ arpa.2025-0232-OA)
Metrics
1 Record Views
Details
- Title
- Accuracy of Foundation Artificial Intelligence Models for Hepatic Macrovesicular Steatosis Quantification in Frozen Sections
- Creators
- Shunsuke Koga - Hospital of the University of PennsylvaniaAnjani Guda - Drexel University, College of MedicineYujie Wang - University of MiamiAarush Sahni - Hospital of the University of PennsylvaniaJiahui WuAlyssa Rosen - Hospital of the University of PennsylvaniaJaxson Nield - Hospital of the University of PennsylvaniaNilan Nandish - Hospital of the University of PennsylvaniaKrunal Patel - Hospital of the University of PennsylvaniaHaviva Goldman - Drexel University, Neurobiology and AnatomyChamith S. Rajapakse - University of PennsylvaniaSelemon Walle - Hospital of the University of PennsylvaniaKristen Stashek - Hospital of the University of PennsylvaniaRashmi Tondon - Hospital of the University of PennsylvaniaZahra Alipour - Hospital of the University of Pennsylvania
- Publication Details
- Archives of pathology & laboratory medicine (1976), v 150(5), pp 374-380
- Publisher
- Coll Amer Pathologists
- Number of pages
- 7
- Resource Type
- Journal article
- Language
- English
- Academic Unit
- Neurobiology and Anatomy; College of Medicine
- Web of Science ID
- WOS:001765764200005
- Scopus ID
- 2-s2.0-105037132727
- Other Identifier
- 991022185375604721