Journal of hepatologyJClub
Validation of a pan-ELastography Machine-learning (ELM) score to predict clinically significant portal hypertension in compensated advanced chronic liver disease.
Journal of hepatology · · Validation Study · Open access
Giuffrè, Kresevic + more
Abstract ↗AI summary
The abstract is read at the publisher; the summary is JClub's.
A pan-elastography machine-learning ELM score predicts clinically significant portal hypertension in compensated advanced chronic liver disease with external AUC 0.91 and reduced gray-zone classification to 12.3%.
- Why it matters: Current Baveno VII CSPH criteria leave 40–50% of compensated advanced chronic liver disease patients in an indeterminate gray zone, limiting safe non-invasive decision-making across elastography types.
- What they did: The model was trained on 943 patients and internally validated on 150, then externally validated in 342 patients using paired hepatic venous pressure gradient, liver stiffness, spleen stiffness, and clinical variables.
- The result: Using cut-offs ≤0.45 and ≥0.60, the ELM score achieved NPV 0.91 (95% CI 0.84–0.95) and PPV 0.96 (95% CI 0.92–0.98), reducing gray zone versus Baveno VII (47.9%) and Baveno-SSM criteria.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.
- Open access
- 1 cites