Multicenter Study to Develop and Validate an AI System for Detecting Small Bowel Damage in Celiac Disease and Other Enteropathies

What this Trial Is About

This research aims to develop and validate an Artificial Intelligence AI system to detect and characterize small bowel mucosal atrophy and other lesions in patients with Celiac Disease CD and Non-Celiac Enteropathies NCEs. These conditions can be challenging to diagnose due to patchy lesions and difficulties in differentiating from other enteropathies. The study addresses the subjectivity and variability in current endoscopic evaluations by using AI to assist in diagnosis across multiple centers. The study is observational and proceeds in two phases over 36 months. The first phase involves training AI algorithms retrospectively on anonymized endoscopic images and videos to identify mucosal atrophy and other pathological findings like ulcers and polyps. The second phase prospectively enrolls adult patients undergoing small bowel endoscopy to validate the AI systems diagnostic accuracy compared to standard human assessment. The AI analysis is performed offline on recorded anonymized videos. Participants include adults undergoing small bowel endoscopy for suspected or confirmed CD, NCEs, or other small bowel lesions, as well as controls without enteropathy. The study collects endoscopic imaging data, compares AI-assisted diagnosis to traditional evaluation, and measures diagnostic performance and inter-observer agreement. The primary outcome is the diagnostic accuracy of the AI system throughout the 36-month study period, with secondary outcomes assessing the difference between AI-assisted and non-assisted diagnosis and consistency among observers.

AI System for Detection and Characterization of Chronic Enteropathies