Clinical Validity Verification on Diagnostic Assistance for Presence and Severity of Colorectal Cancer and Polyps Using Artificial Intelligence-based Medical Device Software Endoscopic Imaging Assistance

This completed retrospective study evaluated the clinical performance of ENAD-CADx-01, an artificial intelligence-based medical device software intended to assist healthcare professionals in characterizing colorectal lesions in colonoscopy images. Previously collected colonoscopy images and corresponding histopathological results from Asan Medical Center were analyzed. The performance of ENAD-CADx-01 was evaluated based on its sensitivity, specificity, and overall accuracy in classifying colorectal lesions as neoplasm, hyperplastic polyp, or sessile serrated lesion. No participants were prospectively enrolled, and no study-related intervention was performed.