Colorectal cancer remains a common malignancy and a leading cause of cancer-related mortality in the United States. Colonoscopy is widely utilized for screening, with the goal of early detection of malignancy, as well as identification and removal of pre-cancerous lesions. Screening strategies are guided by individual risk factors, including family history, the presence of hereditary conditions associated with increased colorectal cancer risk, and findings from prior colonoscopic examinations, including polyp number, size, and histologic characteristics. These factors inform recommendations on the appropriate timing for initiation and surveillance intervals. Endoscopic resection is generally preferred for the removal of adenomatous polyps when feasible, while more advanced lesions may require surgical intervention or specialized endoscopic techniques. Despite advances in technique and operator experience, lesions may still be missed during colonoscopy. To address this limitation and increase the yield of colonoscopies detecting lesions, artificial intelligence (AI)-based tools have been increasingly incorporated to enhance polyp detection and improve overall procedural quality. These technologies aim to support endoscopists in identifying subtle lesions, thus increasing malignancy and polyp detection rates, and thereby potentially contributing to more effective colorectal cancer prevention, lowering healthcare costs, and ultimately improving overall patient outcomes.
Azizian et al. (Mon,) studied this question.