AI Performs Well for Endoscopic Assessment of Crohn Disease

TUESDAY, Sept. 9, 2025 (HealthDay News) -- Artificial intelligence shows good performance for automating ulceration and mucosal injury quantitation with Crohn disease, according to a study published online Aug. 18 in Clinical Gastroenterology and Hepatology.Lingrui Cai, from the University of Michigan in Ann Arbor, and colleagues compared performance of a computer vision model to standard instruments like the Simple Endoscopic Score for Crohn's Disease (SES-CD) to quantify disease activity. The analysis included 4,487 still images from endoscopic videos.The researchers found that ulcer semantic segmentation models matched the performance of gastroenterologist annotators (Dice similarity coefficient, 0.591 versus 0.462), with neither performing better on qualitative review of disagreements. There was a high correlation between computer vision endoscopic (CVE) assessment measures and SES-CD scores (r = 0.73 to 0.85), but there was expected poor correlation with the degree of stenosis (r = 0.12 to 0.21). CVE measures had greater effect size than SES-CD (g = 0.416 versus 0.290) for separating end-of-study clinical remission status (CVE ulcer measurements: 28.4 versus 52.3; SES-CD: 6.3 versus 9.0). A validation cohort performance of CVE was similar."This paper is really the first step, and I hope that it gets the entire field thinking of better ways to quantify IBD that more completely describe an individual patient," senior author Ryan W. Stidham, M.D., also from University of Michigan, said in a statement. "It's very early, but computer vision tools like this may provide an important component for the future of automated care, where artificial intelligence and experts work together in treating patients."Several authors disclosed ties to relevant organizations.Abstract/Full Text (subscription or payment may be required).Sign up for our weekly HealthDay newsletter

AI Performs Well for Endoscopic Assessment of Crohn Disease

Related Stories