Evaluating Real-time AI-assisted Endocytoscopy to Diagnose Small Colorectal Polyps

What this Trial Is About

Colorectal cancer is a common and serious cancer worldwide, with colonoscopy being the preferred method for screening. Early detection of colorectal neoplastic lesions can reduce disease impact, but many detected polyps are small or non-neoplastic and do not require removal or pathological examination. This study focuses on evaluating an artificial intelligence AI assisted endocytoscopy system designed to help diagnose the nature of small colorectal polyps in real-time, aiming to improve the accuracy of diagnosis and reduce unnecessary procedures. The AI system uses ultra-high magnification endoscopy combined with chemical staining and imaging to observe cellular details of colorectal lesions. This allows endoscopists to make real-time biopsy decisions during colonoscopy. The study intends to verify the diagnostic accuracy of this AI-assisted method for small colorectal polyps, supporting strategies to excise only necessary lesions and avoid unnecessary pathology. Participants with colorectal lesions will be observed using this AI-assisted endocytoscopy. Researchers will measure diagnostic accuracy through negative predictive value, sensitivity, specificity, and overall accuracy. The study runs until December 31, 2025, and aims to gather clinical data to support wider use of AI in endoscopic diagnosis of colorectal diminutive polyps, potentially reducing costs and improving patient care.

Clinical Study on the Accuracy of Real-time AI-assisted Endocytoscopy in the Diagnosis of Colorectal Diminutive Polyps