Truace
Subscribe

FDA and EMA qualified AI models for liver disease trials; a review lists more possible uses

Study title: How to apply artificial intelligence (AI) to facilitate and enhance MASH trials

Published September 16, 2026, this narrative review examines how AI is being applied to metabolic dysfunction-associated steatohepatitis trials, a field where only resmetirom and semaglutide had received conditional approval at the time of writing.

Read the study abstract

Introduction Metabolic dysfunction-associated steatohepatitis (MASH) is one of the most common liver diseases, but thus far only two drugs (resmetirom and semaglutide) have received conditional approval for use. Successful MASH trials are thus highly desired to fill the therapeutic and evidence gap. Concurrently, AI has powered advances in multiple areas of medicine. We consider in this review the potential applications of AI to enhance MASH trials. Methods We conduct a narrative review of AI uses in MASH, with…

Truvace record1 min read1 sourceEvidence: Emerging · 30/100 — Commentary, opinion, or narrative review · 1 independent sourcePeer-reviewedTRV-2026-1134 · cite or check it
How to apply artificial intelligence (AI) to facilitate and enhance MASH trials
"Respiratory bronchiolitis" by Pulmonary Pathology is licensed under CC BY-SA 2.0. To view a copy of this license, visit https://creativecommons.org/licenses/by-sa/2.0/.

In brief

Regulatory qualification of AI pathology models for endpoint determination signals a shift in how liver disease trials are run, but the review remains largely descriptive of potential uses, leaving open how broadly these tools improve enrollment, retention, and trial validity in practice.

Main points

  1. MASH is described as one of the most common liver diseases with only two drugs having received conditional approval.
  2. FDA and EMA have qualified specific AI models for endpoint determination in MASH trials.
  3. Review also covers AI risk models for enrollment and large language models for patient interactions.

The gain

AI-based digital pathology models have been qualified by regulators for endpoint determination in MASH trials, with additional AI risk models and large language models applied to patient enrollment, recruitment and retention.

The full story

The review describes digital pathology and AI technologies for MASH grading as a mature application, noting regulatory qualification for trial endpoints.

It further details contemporary AI risk models to facilitate patient enrollment and large language models to support recruitment and retention, drawing lessons from other medical domains and current guidelines.

Sources

  1. Peer-reviewedHepatology International2026-09-16

The debate