AI pathology model outperforms human pathologists in evaluating NASH histological features
The use of a machine-learning (ML) pathology model, in comparison to human pathologists, demonstrated superior performance in evaluating histological features of non-alcoholic steatohepatitis (NASH), according to a study.
The post hoc analysis involved data from a subset of patients (n = 251) with biopsy-confirmed NASH and fibrosis stage F1-F3 from a 72-week randomized placebo-controlled trial of once-daily subcutaneous semaglutide. Biopsies at baseline and week 72 were independently read by two pathologists. Digitized biopsy slides were then assessed by PathAI’s NASH ML models to quantify changes in fibrosis, steatosis, inflammation, and hepatocyte ballooning.
The results found that both pathologists and ML-derived categorical assessments detected a significantly greater percentage of patients achieving the primary endpoint of NASH resolution without worsening of fibrosis with semaglutide 0.4 mg versus placebo. Specifically, the pathologist group showed a percentage of 58.5% compared to 22.0% in the placebo group (P < 0.0001), while the ML group exhibited 36.9% versus 11.9% (P = 0.0015). Additionally, both methods detected a higher but nonsignificant percentage of patients on semaglutide 0.4 mg versus placebo achieving the secondary endpoint of liver fibrosis improvement without NASH worsening.
The most significant finding was the ML-based continuous scores, which detected a treatment-induced antifibrotic effect not measured by conventional histopathology. The continuous scores showed a quantitative reduction in fibrosis with semaglutide 0.4 mg versus placebo (P = 0.0099), a result that was not evident in either pathologist or ML categorical assessments.
Reference
Ratziu V, Francque S, Behling CA, et al. Artificial intelligence scoring of liver biopsies in a phase II trial of semaglutide in nonalcoholic steatohepatitis. Hepatology. 2023;doi: 10.1097/HEP.0000000000000723. Epub ahead of print. PMID: 38112484.
