Endoscopic AI-assisted Decision-making for Ileocecal Ulcer Diagnosis: A Multi-Reader Multi-Case (MRMC) Randomized Controlled Trial

INTERVENTION: Gold Standard:Final confirmed clinical diagnosis is used as the reference standard, determined by a combined diagnostic approach: Crohn’s disease (CD): diagnosed per Chinese Guidelines for Crohn’s Disease (2023, Guangzhou) and ECCO consensus; Intestinal Behçet’s disease (BD): diagnosed per Chinese Consensus on Diagnosis and Treatment of Intestinal Behçet’s Disease; Intestinal tuberculosis (ITB): meeting any of the following: positive Mycobacterium tuberculosis culture, typical AFB/caseating granuloma, positive GeneXpert/MTB PCR, or significant clinical improvement after 2–6 weeks of anti‐TB treatment with endoscopic improvement at 2–3 months. All cases were confirmed by at least 12 months of follow‐up, and equivocal cases were excluded.;Inde Xtest:Endoscopic AI‐assisted model for differential diagnosis of ileocecal ulcers (multimodal deep learning model for classification of CD/ITB/intestinal BD); CONDITION: Crohn' PRIMARY OUTCOME: Area Under ROC Curve (AUC); INCLUSION CRITERIA: 1.Aged 20–45 years; 2.Gastroenterologists or novice endoscopists with knowledge of digestive tract anatomy (3–5 years of independent endoscopic experience); 3.Signed informed consent.