Colorectal cancer (CRC) is a major public health concern, emphasizing the importance of early detection and refining current screening methods. This systematic review aimed to evaluate the advantages and drawbacks of AI-based computer-aided detection (CADe) systems compared to standard colonoscopy. The review highlights the global prevalence and mortality rates of CRC and the challenges of undetected adenomas and polyps. It summarizes findings from 13 randomized controlled trials (RCTs) from 2019 to 2023, assessing the impact of CADe on polyp and adenoma detection. The results consistently indicate that CADe outperforms traditional colonoscopy in terms of adenoma detection rate (ADR) and polyp detection rate (PDR), especially for small and flat lesions that are often missed.
Despite some limitations in the studies, such as potential biases and geographic constraints, the review concludes that AI-assisted colonoscopy can decrease missed lesion rates and enhance CRC diagnosis. Collaboration between experts and clinicians is essential for effective implementation. The review suggests that more double-blinded trials and studies among diverse populations from various countries are needed to further validate and expand these findings.