As AI continues to reshape the future of healthcare, a significant milestone has been achieved. The European Medicines Agency’s (EMA) human medicines committee (CHMP) has granted the first-ever Qualification Opinion (QO) for an AI-based development tool. The approved tool, known as AIM-NASH, is designed to assist pathologists in analyzing liver biopsy scans to determine the severity of MASH (metabolic dysfunction-associated steatohepatitis)
MASH is a health condition that causes fat accumulation in the liver leading to tissue inflammation & scarring. People who are more susceptible to this disease are more likely to struggle from obesity, type-2 diabetes, high blood pressure or abnormal cholesterol levels.
Traditionally, MASH diagnosis in clinical trials depended on liver biopsies reviewed by multiple pathologists. However, this process can lead to inconsistent results due to human interpretation differences. AIM-NASH aims to reduce this variability by providing more consistent, repeatable, and objective assessments of liver inflammation and fibrosis.
The AIM-NASH system has been developed through employing a machine learning model that has been trained using 100,000 annotations from 59 pathologists who assessed over 5,000 liver biopsies across nine large clinical trials.
Typically, the machine learning model is locked & is not eligible for modification. However, the CHMP encourages future model optimization noting that any major updates would require a reassessment.
All activities utilizing Artificial Intelligence used by the EMA is based on multiannual AI workplan by the EMA & The Heads of Medicines Agencies, which works to provide safe and responsible use of AI across the European medicines regulatory network.
