Real-time, computer-aided, detection-assisted colonoscopy eliminates differences in adenoma detection rate between...

Area of Science:

• Gastroenterology
• Medical Technology
• Endoscopy

Background:

• Adenoma detection rate (ADR) is a key quality metric for colonoscopy screening.
• Artificial intelligence (AI) has shown promise in improving ADR and reducing adenoma miss rates (AMR).
• The impact of AI on the performance of trainee endoscopists (TEs) remains understudied.

Purpose of the Study:

• To evaluate if AI assistance can bridge the performance gap in ADR and AMR between TEs and experienced endoscopists (EEs).
• To assess AI's potential to standardize colonoscopy quality during training.

Main Methods:

• A prospective observational study involving 45 patients undergoing screening colonoscopy.
• Trainee endoscopists (TEs) performed colonoscopies with AI assistance, followed by same-day tandem examinations by experienced endoscopists (EEs) unaware of TE findings.
• Key metrics included ADR, AMR, adenoma per colonoscopy (APC), polyp detection rate (PDR), polyp per colonoscopy (PPC), and polyp miss rate (PMR).

Main Results:

• AI-assisted TEs achieved an ADR of 38% and APC of 0.93, comparable to EEs (ADR 40%, APC 1.07).
• No significant differences were observed in ADR, APC, PDR, or PPC between AI-assisted TEs and EEs (P > 0.05).
• AMR and PMR for AI-assisted TEs were 12.5% and 13%, respectively, with no significant differences compared to EEs.

Conclusions:

• AI demonstrates a potential impact on the quality training of endoscopists.
• AI assistance may help TEs achieve performance levels similar to EEs.
• Future applications of AI could enhance screening colonoscopy efficacy by reducing interval and missed cancers.