Machine learning algorithm improves the detection of NASH (NAS-based) and at-risk NASH - A development and validation study

Aug 5, 2026

Upload by Anda

Detecting NASH remains challenging, and at-risk NASH (steatohepatitis and F≥2) often progresses, making accurate identification important for drug development. A machine learning approach was developed and validated to improve detection of NASH using NAS-based definitions and to identify individuals with at-risk NASH. The study evaluates model performance in development and validation settings, aiming to enhance diagnostic reliability compared with existing limitations.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages14

Reading Time35 min

AccessFree to download

Abstract

Background and Aims

Methods (development and validation)

Results (model performance)

Conclusion

Downloaded from [http://journals.lww.com/hep by BhDMf5ePHKav1zEoum1tQfN4a](http://journals.lww.com/hep by BhDMf5ePHKav1zEoum1tQfN4a)+kJLhEZgbsIHo4XMi0hCywCX1AWn YQp/ I l Qr HD3i 3 D0OdRyi 7TvS Fl4Cf3VC 1y0abggQZXdtwnfKZBYtws= on 07/28/2023 DOI: 10.1097/HEP.0000000000000364 O R IG INAL ART ICL E Machine learning algorithm improves the detection of NASH (NAS-based) and at-risk NASH: A development and validation study Jenny Lee1 | Max Westphal2 | Yasaman Vali 1 | Jerome Boursier3 | Salvatorre Petta4 | Rachel Ostroff5 | Leigh Alexander5 | Yu Chen6 | Celine Fournier7 | Andreas Geier8 | Sven Francque9 | Kristy Wonders10 | Dina Tiniakos10, 11 | Pierre Bedossa10 | Mike Allison 12 | Georgios Papatheodoridis13 | Helena Cortez-Pinto14 | Raluca Pais 15 | Jean-Francois Dufour16 | Diana Julie Leeming 17 | Stephen Harrison 18 | Jeremy Cobbold 18 | Adriaan G. Holleboom 19 | Hannele Yki-Järvinen20 | Javier Crespo21 | Mattias Ekstedt22 | Guruprasad P. Aithal23 | Elisabetta Bugianesi24 | Manuel Romero-Gomez25 | Richard Torstenson26 | Morten Karsdal 17 | Carla Yunis27 | Jörn M. Schattenberg28 | Detlef Schuppan29,30 | Vlad Ratziu31 | Clifford Brass32 | Kevin Duffin6 | Koos Zwinderman 1 | Michael Pavlides33 | Quentin M. Anstee9,34 | Patrick M. Bossuyt 1 | on behalf of the LITMUS investigators 1Department of Epidemiology and Data Science, Amsterdam UMC, Amsterdam, the Netherlands 2Fraunhofer Institute for Digital Medicine MEVIS, Bremen, Germany 3Department of Hepatology, Angers University Hospital, Angers, France 4Section of Gastroenterology and Hepatology, Promozione della Salute, Materno-Infantile, di Medicina Interna e Specialistica di Eccellenza, Department, University of Palermo, Palermo, Italy 5SomaLogic Inc, Boulder, Colorado, USA 6Lilly Research Laboratories, Eli Lilly and Company Ltd (LLY), Indianapolis, Indiana, USA 7Echosens, 6 rue Ferrus, Paris, France 8Division of Hepatology, Department of Medicine II, Wurzburg University Hospital, Wurzburg, Germany 9Department of Gastroenterology Hepatology, Antwerp University Hospital, and Laboratory of Experimental Medicine and Paediatrics, University of Antwerp, Belgium 10Translational and Clinical Research Institute, Faculty of Medical Sciences, Newcastle University, Newcastle upon Tyne, UK 11Department of Pathology, Aretaieion Hospital, national and Kapodistrian University of Athens, Athens, Greece 12Liver Unit, Department of Medicine, Cambridge NIHR Biomedical Research Centre, Cambridge University NHS Foundation Trust, CB2 0QQ, Cambridge, UK 13Gastroenterology Department, National and Kapodistrian University of Athens, General Hospital of Athens “Laiko”, Athens, Greece 14Clínica Universitária de Gastrenterologia, Faculdade de Medicina, Universidade de Lisboa, Portugal 15Assistance Publique-Hôpitaux de Paris, hôpital Pitié Salpêtrière, Sorbonne University, ICAN (Institute of Cardiometabolism and Nutrition), Paris, France 16Hepatology, Department of Biomedical Research, University of Bern, Bern, Switzerland 17Nordic Bioscience A/S, Herlev, Denmark 18Department of Gastroenterology and Hepatology, Oxford NIHR Biomedical Research Centre, John Radcliffe Hospital, Oxford, UK 19Department of Internal and Vascular Medicine, Amsterdam University Medical Centres, location AMC, Amsterdam, the Netherlands [258](258 | www.hepjournal.com)[ |](258 | www.hepjournal.com)[ www.hepjournal.com](258 | www.hepjournal.com) Hepatology. 2023;78:258–271 Downloaded from [http://journals.lww.com/hep by BhDMf5ePHKav1zEoum1tQfN4a](http://journals.lww.com/hep by BhDMf5ePHKav1zEoum1tQfN4a)+kJLhEZgbsIHo4XMi0hCywCX1AWn YQp/ I l Qr HD3i 3 D0OdRyi 7TvS Fl4Cf3VC 1y0abggQZXdtwnfKZBYtws= on 07/28/2023 20Department of Medicine, University of Helsinki and Helsinki University Hospital, Finland; Minerva Foundation Institute for Medical Research, Helsinki, Finland 21Department of Gastroenterology and Hepatology, University Hospital Marques de Valdecilla. Resea

FAQs

Why is detecting NASH considered challenging, and why is at-risk NASH important?

NASH detection remains challenging, while at-risk NASH (steatohepatitis with F≥2) tends to progress. Accurate identification supports drug development and timely intervention.

What does the study develop and validate?

The study develops and validates a machine learning algorithm intended to improve detection of NASH using NAS-based criteria and to identify at-risk NASH.

How is the work structured to assess the algorithm’s performance?

Performance is evaluated across development and validation settings to measure how well the model generalizes and supports reliable detection.

Aug 5, 2026

Upload by Anda