Computer-assisted adenoma detection colonoscopy with an ENDO-AID artificial intelligence system and Endocuff- versus...

Area of Science:

• Gastroenterology
• Endoscopy
• Cancer Prevention

Background:

• Computer-assisted adenoma detection (CADe) and Endocuff-Vision (EV) individually enhance adenoma detection rates (ADR) in colonoscopy.
• Limited evidence exists on the combined efficacy of CADe and EV versus EV alone.

Purpose of the Study:

• To compare the adenoma detection rate (ADR) of combined CADe (ENDO-AID) and EV versus EV alone.
• To evaluate secondary outcomes including advanced ADR and adenomas per colonoscopy (APC).

Main Methods:

• A randomized controlled trial involving 696 patients comparing CADe plus EV versus EV alone.
• Primary outcome: ADR. Secondary outcomes: advanced ADR, APC.
• Intention-to-treat (ITT) and per-protocol (PP) analyses were performed.

Main Results:

• No significant difference in overall ADR between the combination and EV alone groups in ITT (54.6% vs. 51.1%) or PP (58.7% vs. 56.6%) analyses.
• Significantly higher ADR in the proximal colon for the combination group (39.6% vs. 30%, P=0.015).
• Significantly higher APC in the combination group (mean 1.7 vs. 1.31, P=0.029), driven by increased detection of flat diminutive adenomas.

Conclusions:

• Colonoscopy with combined ENDO-AID (OIP-1) and EV does not significantly increase the overall ADR compared to EV alone.
• The combination demonstrates potential benefits in detecting proximal and diminutive flat adenomas.