Computer-assisted adenoma detection colonoscopy with an ENDO-AID artificial intelligence system and Endocuff- versus Endocuff-assisted colonoscopy: A randomized controlled trial.

BACKGROUND

Compared with conventional colonoscopy, computer-assisted adenoma detection (CADe) colonoscopy and Endocuff-Vision (EV) have been shown to increase the adenoma detection rate (ADR). However, the evidence for whether their combination further increases the ADR compared with that of EV alone is limited. The aim was to compare the combination of CADe and EV, with EV alone.

METHODS

In all, 696 patients were randomized 1:1 to receive either CADe (ENDO-AID [OIP-1]) plus EV or EV alone at a tertiary care hospital. The primary outcome was the ADR. The secondary outcomes included the advanced ADR and the number of adenomas per colonoscopy (APC). Both intention-to-treat (ITT) and per-protocol (PP) analyses were conducted for the primary outcome.

RESULTS

In the ITT analysis, the ADR was 54.6% and 51.1% in the combination and EV groups, respectively (OR 1.15, 95% CI (0.85-1.55)). In the PP analysis, the ADRs were 58.7% and 56.6%, respectively (OR 1.09, 95% CI (0.79-1.52)), for the combination and EV groups. The ADR in the proximal colon was significantly greater in the combination group than in the EV group (39.6% vs. 30%; P=0.015). The APC value was significantly greater in the combination group than in the EV group (mean: 1.7, SD: 2.34 vs. mean: 1.31, SD: 1.92; P=0.029), which was primarily due to increased detection of flat diminutive adenomas.

CONCLUSIONS

Compared with EV alone, colonoscopy assisted by both ENDO-AID (OIP-1) and EV does not significantly increase the ADR. However, this combination may be beneficial for the detection of proximal and diminutive flat adenomas (NCT05141773).