Integration of Endocuff-Assisted and Computer-Aided Colonoscopy: A Meta-Analysis of Randomized Controlled Trials

Colorectal cancer (CRC) is a leading cause of cancer-related deaths, with missed lesions during colonoscopy contributing to increased mortality. AI-assisted computer-aided detection (CADe) systems help reduce adenoma miss rates, and their integration with mucosal exposure devices like EndoCuff vision may enhance detection. This meta-analysis assesses the effectiveness of combining CADe and EndoCuff vision-assisted colonoscopy (EAC) compared to standard colonoscopy or CADe alone.

A comprehensive literature search was conducted in MEDLINE, Embase, and clinicaltrials.gov up to November 2024. Only randomized controlled trials (RCTs) comparing CADe+EAC with CADe alone or standard colonoscopy and reporting adenoma detection rate (ADR) were included. Statistical analysis was performed using R version 4.4.0, with mean differences (MDs) and risk ratios (RRs) reported with 95% confidence intervals (CIs).

Four RCTs with 3333 participants were included. Among them, 1228 underwent CADe+EAC, while 554 received standard colonoscopy and 1236 underwent CADe alone. CADe+EAC significantly improved ADR (RR: 1.36; 95% CI: 1.19-1.56), advanced ADR (RR: 1.62; 95% CI: 1.16-2.25), sessile serrated lesion detection rate (RR: 1.95; 95% CI: 1.46-2.61), and mean adenomas per colonoscopy (APC) (MD: 0.50; 95% CI: 0.49-0.52) compared to standard colonoscopy. Compared to CADe alone, CADe+EAC further improved ADR (RR: 1.13; 95% CI: 1.04-1.22) and mean APC (MD: 0.15; 95% CI: 0.07-0.24). However, withdrawal and insertion times were significantly lower in the CADe+EAC group.

The combination of CADe and EAC enhances adenoma detection compared to standard colonoscopy and CADe alone. Further multi-center trials with diverse AI models are needed to confirm its effectiveness.