The EMA has qualified AIM-NASH, the first AI tool for diagnosing inflammatory liver disease (MASH) using biopsy samples. This innovation enhances clinical trial reliability by minimizing variability in assessing disease severity, promising quicker access to effective treatments. The tool is locked for consistency, with comprehensive AI management under EMA’s oversight for safe use.
The European Medicines Agency (EMA) has granted its first Qualification Opinion (QO) for an artificial intelligence tool designed to diagnose inflammatory liver disease. This innovative tool, named AIM-NASH, assists pathologists in analyzing liver biopsy scans to determine the severity of metabolic dysfunction associated steatohepatitis (MASH). MASH, previously referred to as non-alcoholic steatohepatitis, involves fat accumulation in the liver leading to severe complications without significant alcohol consumption.
AIM-NASH aims to enhance the precision and efficiency of clinical trials for MASH treatments, reducing variability in disease activity assessments such as inflammation and fibrosis. By using this tool, researchers can achieve more reliable evidence of new treatments’ benefits, even with fewer trial participants, thus accelerating access to effective therapies.
Traditionally, the efficacy of new MASH treatments has hinged on liver biopsies—considered the gold standard for assessing liver inflammation and damage. However, variability in trial outcomes presents a significant challenge, as pathologists may disagree on the severity of liver conditions. AIM-NASH addresses this by providing consistent biopsy readings validated by expert pathologists, demonstrating greater reliability than traditional methods, which often require consensus from multiple experts.
The AIM-NASH tool employs advanced machine learning techniques, having been trained on over 100,000 annotations from 59 pathologists across more than 5,000 liver biopsies in nine extensive clinical trials. This ‘locked’ tool ensures that the machine learning model remains unaltered, with the EMA stating that any significant modifications would necessitate re-qualification.
EMA’s AI initiatives are guided by a comprehensive workplan, coordinated with the Heads of Medicines Agencies, dedicated to promoting the safe and responsible integration of AI technologies in the European medicines regulatory environment, ensuring rigorous standards are met across the board.
The EMA’s Qualification Opinion for the AIM-NASH AI tool marks a significant step forward in diagnosing and assessing MASH. By providing more consistent biopsy analysis, this innovative tool is set to enhance the reliability of clinical trials, streamlining the development of new treatments. This advancement not only highlights the potential of AI in healthcare but also reinforces the EMA’s commitment to integrating advanced technologies responsibly in medical regulation.
Original Source: www.ema.europa.eu
