<![CDATA[Colorectal cancer remains one of the leading causes of cancer-related morbidity and mortality worldwide. Colonoscopy plays a central role in colorectal cancer screening, and the adenoma detection rate (ADR) is a key quality indicator associated with reduced colorectal cancer risk. Recently, artificial intelligence (AI)-based computer-aided detection (CADe) systems have been developed to assist endoscopists in identifying colorectal lesions during colonoscopy. This research aimed to evaluate the effectiveness of artificial intelligence–assisted colonoscopy in improving adenoma detection compared with conventional colonoscopy. A systematic review and meta-analysis were conducted following PRISMA guidelines. Literature searches were performed in PubMed, the Cochrane Library, and ScienceDirect to identify comparative studies evaluating CADe-assisted colonoscopy versus conventional colonoscopy. The primary outcome was adenoma detection rate. Pooled risk ratios (RRs) with 95% confidence intervals (CIs) were calculated using a random-effects model. Four studies comprising 10,483 colonoscopy procedures were included in the meta-analysis. The pooled results demonstrated that AI-assisted colonoscopy significantly increased adenoma detection compared with conventional colonoscopy (RR = 1.19; 95% CI = 1.07–1.32; p = 0.002). Moderate heterogeneity was observed among the included studies (I² = 68%). Notably, although overall ADR improved substantially, several studies reported limited or no significant improvement in the detection of advanced colorectal neoplasia, suggesting that the benefit may have been predominantly driven by increased detection of diminutive adenomas. Artificial intelligence–assisted colonoscopy using CADe systems significantly improved adenoma detection rates compared with conventional colonoscopy. However, further large-scale randomized studies are needed to determine whether these improvements translate into clinically meaningful long-term reductions in colorectal cancer incidence and mortality.]]>
Copyrights © 2026
