Real-time use of a computer-aided system for polyp detection during colonoscopy, an ambispective study

Area of Science:

• Gastroenterology
• Medical Imaging
• Endoscopic Technology

Background:

• Colonoscopy is a key procedure for colorectal cancer screening and diagnosis.
• Improving polyp detection is crucial for enhancing colonoscopy efficacy and patient outcomes.
• Computer-aided detection (CADe) systems offer potential to augment the performance of endoscopists.

Purpose of the Study:

• To evaluate the effectiveness of a real-time computer-aided detection (CADe) system.
• To assess the impact of CADe on polyp per colonoscopy (PPC) and adenoma detection rate (ADR).
• To compare polyp detection metrics between standard colonoscopy and CADe-assisted colonoscopy.

Main Methods:

• Retrospective analysis of 85 colonoscopy videos with and without CADe system marking.
• Prospective evaluation of 128 patients undergoing standard or CADe-assisted colonoscopy alternately.
• Comparison of polyp detection rate (PDR), adenoma detection rate (ADR), PPC, and APC between groups.

Main Results:

• Marked videos showed a higher number of polyps (88 vs 73) and mean PPC (1.04 vs 0.86).
• Prospective CADe colonoscopy resulted in a higher mean PPC (1.66 vs 1.13) and significantly increased PDR (78.1% vs 56.3%).
• The proportion of polyps detected per colonoscopy increased by 20.5% with CADe.

Conclusions:

• Real-time CADe systems significantly enhance polyp detection rate (PDR) and polyps per colonoscopy (PPC).
• The CADe system demonstrates effectiveness in improving polyp detection, particularly in settings with high detection rates.
• This technology represents a valuable tool for improving the quality and efficacy of colonoscopic examinations.