Identify under-and un-diagnosed NASH patients with deep clinical data and advanced AI

Challenge

Identification of both probable and confirmed non-alcoholic fatty liver disease (NAFLD) and NASH is hampered by asymptomatic disease, under-diagnosis in routine care, lack of condition-specific treatments and poor billing code specificity.

Solution

Leveraging deep clinical data on millions of patients, OM1:

• Used AI to identify patients who were fast progressors from NAFLD to NASH
• Developed AI algorithms to automatically learn the characteristics of confirmed NASH cases
• Characterized the likelihood that a patient who is not a confirmed NASH patient is in fact a NASH patient and apply it to a sample database of over 44 million at risk patients

Key Takeaways

• The OM1 NASH Score is a highly predictive score that indicates the probability that a patient has NASH.
• Our score can help identify large populations of under-diagnosed patients with high confidence.
• This approach can potentially be the foundation for early detection to a large number of under-diagnosed and/
  or poorly coded conditions using AI.