Effectiveness of the GI Genius Computer-Aided Detection System Versus Standard Colonoscopy: A Systematic Review and Meta-Analysis of Randomized Controlled Trials.

Abstract

Colorectal cancer prevention relies on high-quality colonoscopy, yet clinically relevant lesions are still missed. GI Genius is an FDA-cleared computer-aided detection (CADe) system that flags suspected polyps in real time. We conducted a Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA)-aligned systematic review and meta-analysis of randomized controlled trials in adults comparing GI Genius-assisted versus standard colonoscopy. PubMed, Ovid Embase, and Cochrane Central Register of Controlled Trials (CENTRAL) were searched from the earliest available records in each database through August 25, 2025, with citation chasing. Seven trials (n = 9,639) met the eligibility criteria. Random-effects meta-analyses showed that GI Genius increased the adenoma detection rate (ADR) compared with standard colonoscopy (RR 1.12, 95% CI 1.03-1.22; I²=64%). Secondary outcomes also favored GI Genius: polyp detection rate (PDR; two trials, n = 5,225; RR 1.05, 95% CI 1.01-1.10; I² = 0%), sessile serrated lesion detection rate (SSLDR; four trials, n = 7,013; RR 1.27, 95% CI 1.11-1.47; I² = 11%), and lesion counts (adenomas per colonoscopy/mean adenomas per patient, six trials, n = 9,253; standardized mean difference (SMD) 0.15, 95% CI 0.09-0.20; I² = 29%; polyps per colonoscopy/mean polyps, two trials, n = 5,228; SMD 0.10, 95% CI 0.05-0.15; I² = 0%). Advanced ADR (three trials, n=6,328) showed no significant difference (RR 1.01, 95% CI 0.90-1.13; I² = 6%). Safety reporting was limited and was not included in the meta-analysis. In summary, adjunctive GI Genius improves clinically meaningful detection, increasing ADR, PDR, SSLDR, and per-procedure lesion counts, without a detectable effect on advanced adenoma detection. These findings support routine activation of GI Genius during eligible colonoscopies, contingent on appropriate team training and consistent response to system prompts. Larger multicenter trials with standardized methods and long-term clinical endpoints are warranted.