AI Shows Promise for Detecting Ulcerative Colitis Healing, But Needs More Consistent Training
-
AI systems show high potential for detecting mucosal healing in ulcerative colitis, with optimal diagnostic performance based on images and videos.
-
AI algorithms replicated expert opinion with high sensitivity and specificity, but moderate-high heterogeneity of data was found.
-
AI software is expected to improve interobserver agreement in assessing mucosal healing compared to human endoscopists.
-
The meta-analysis found AI systems achieved satisfactory performance in evaluating mucosal healing from images and videos.
-
More training consensus and expert-validated databases are needed due to variability in how AI systems are trained currently.