AI-Driven Multimodal Imaging Integration for Diagnosis and Prognostication of Digestive System Diseases

What this Trial Is About

Researchers are evaluating a new noninvasive system that uses artificial intelligence AI to analyze multiple types of imaging, including magnetic resonance enterography MRE and computed tomography enterography CTE, to help diagnose and predict outcomes for digestive diseases. This observational study collects and analyzes retrospective imaging, endoscopic, and clinical data from 21 centers in China to build and improve the AI model. The model will then be tested prospectively in two centers and its ability to locate lesions will be checked in real-world endoscopy settings. The study involves using a virtual endoscopy model to assist in diagnosis by integrating and analyzing multimodal imaging features. The AI system is designed to support diagnosis without any invasive procedures. The study will confirm the models accuracy and effectiveness through retrospective data, prospective validation, and real-world deployment in clinical environments. Participants will contribute data from their imaging and endoscopic exams, with at least one technically adequate CT or MR scan and a high-quality colonoscopy performed within one month of imaging. Researchers will assess the AI models diagnostic performance by measuring the area under the ROC curve AUC over six months. The study includes ongoing monitoring of data quality and imaging accuracy to ensure reliable validation of the AI system.

AI-Driven Multimodal Imaging Integration for Diagnosis and Prognostication of Digestive System Diseases