Particle.news
Download on the App Store

FDA Qualifies First AI Tool for MASH Drug Trials

The decision makes a cloud-based pathology algorithm available for trials to reduce variability in liver-biopsy scoring.

The corporate logo of the U.S. Food and Drug Administration (FDA) is shown in Silver Spring, Maryland, November 4, 2009. REUTERS/Jason Reed/File Photo

Overview

  • The FDA qualified AIM-NASH, described by developer PathAI as AIM‑MASH AI Assist, as the first AI tool for assisting MASH assessment in drug development under its Drug Development Tool program.
  • The qualification makes the system publicly available to sponsors for the agency’s specified context of use in clinical trials.
  • The cloud-based algorithm analyzes digitized liver-biopsy images and produces standardized scores for fat, inflammation and scarring that clinicians then interpret.
  • The agency based its decision on studies showing AI‑assisted reads comparable to those of individual experts, with published data citing improved reproducibility over manual scoring.
  • PathAI reports the tool previously received EMA qualification, positioning it as a regulator‑endorsed option to standardize enrollment and endpoint assessment in Phase 2 and Phase 3 MASH studies.