A comprehensive RCT in screening, surveillance, and diagnostic AI-assisted colonoscopies (ACCENDO-Colo study)

Area of Science:

• Gastroenterology
• Medical Imaging
• Artificial Intelligence in Medicine

Background:

• Adenoma detection rate (ADR) is a critical quality indicator for colonoscopy.
• Computer-aided detection (CADe) systems show variable efficacy across patient groups.
• Evaluating novel CADe systems is essential for improving colon cancer screening.

Purpose of the Study:

• To assess the effectiveness of a new CADe system in enhancing ADR.
• To compare AI-assisted colonoscopy with standard colonoscopy in a multicenter trial.
• To evaluate the impact of CADe on secondary outcomes like polyp and adenoma detection rates.

Main Methods:

• A multicenter randomized control trial (RCT) in Italy involving 1158 patients aged 40-85.
• Patients were randomized to either CADe-assisted or standard colonoscopy (SC).
• Exclusion criteria included inadequate bowel preparation (Boston Bowel Preparation Scale < 2).

Main Results:

• The CADe group demonstrated a significantly higher ADR (50.2%) compared to the SC group (40.5%; p=0.001).
• CADe significantly increased polyp per colonoscopy (PPC) and adenoma per colonoscopy (APC) detection.
• No significant difference in sessile serrated lesion detection rate (SSLDR) was observed between groups.

Conclusions:

• AI-assisted colonoscopy significantly improves ADR in a general patient population.
• The CADe system did not prolong procedure time or increase non-neoplastic lesion detection.
• This technology offers a promising tool for enhancing colonoscopy quality and colorectal cancer prevention.