Machine learning algorithm improves detection of NASH (NAS-based) and at-risk NASH, a development and validation study.
Jenny A LeeMax WestphalYasaman ValiJerome BoursierRachel OstroffLeigh AlexanderYu ChenCeline FournierAndreas GeierSven FrancqueKristy WondersDina TiniakosPierre BedossaMike AllisonGeorgios PapatheodoridisHelena Cortez-PintoRaluca PaisJean-Francois DufourDiana Julie LeemingStephen HarrisonJeremy CobboldAdriaan G HolleboomHannele Yki-JärvinenJavier CrespoMattias EkstedtGuruprasad P AithalElisabetta BugianesiManuel Romero-GomezMorten KarsdalCarla YunisJörn M SchattenbergDetlef SchuppanVlad RatziuClifford BrassKevin DuffinKoos ZwindermanMichael PavlidesQuentin M AnsteePatrick M Bossuytnull nullPublished in: Hepatology (Baltimore, Md.) (2023)
Detection of NASH and at-risk NASH can be improved by constructing independent ML models for each component, using only clinical predictors. Adding biomarkers only improved accuracy for fibrosis.