Randomized controlled trials support the efficacy of both computer-aided detection (CADe) and mucosal exposure devices (MED) in improving adenoma detection, but implementation studies—particularly on their combined use in routine practice—are limited. We evaluated whether combining CADe and MED improves adenoma detection rate (ADR) in real-world colonoscopy.

In this prospective implementation study, adults aged ≥45 years undergoing elective colonoscopy received standard, CADe-, MED-, or CADe+MED-assisted procedures. The primary outcome was ADR; secondary outcomes included advanced adenoma detection rate (AADR), sessile serrated lesion detection rate (SSLDR), adenoma per colonoscopy (APC), polyp detection rate (PDR), polyps per colonoscopy (PPC), and withdrawal time (WT).

We included 2552 patients undergoing standard (n = 1279), CADe-assisted (n = 829), MED-assisted (n = 225), or CADe+MED-assisted (n = 219) colonoscopies. ADR was 55.6% (95% CI [48.7–62.4]) in the CADe+MED group versus 29.3% (95% CI [26.5–32.3], p < 0.001) in the standard group. The CADe+MED group showed significant increases in AADR, SSLDR, APC, PDR, and PPC compared to standard procedures (all p < 0.001). CADe- and MED-alone achieved ADRs of 41.6% (95% CI [38.0–45.3]) and 44.9% (95% CI [37.5–52.6]), respectively (p < 0.866). CADe+MED increased ADR compared with CADe alone (p = 0.002). Mean WT was shorter for standard procedures (9.9 min) than for assisted ones (11.9–13.6 min, all p < 0.001).

Combining CADe and MED significantly improves ADR, AADR, SSLDR, and other key performance metrics versus standard colonoscopy. These data suggest a synergistic effect wherein optimized mucosal exposure enables CADe to realize its full detection potential for aiding routine colonoscopies.