AI and Polyp Detection During Colonoscopy

Area of Science:

• Gastroenterology and Medical Imaging
• Artificial Intelligence in Medicine

Background:

• Effective colonoscopy is crucial for colorectal cancer (CRC) prevention, but adenoma detection rates (ADRs) and missed lesions present significant challenges.
• Artificial intelligence (AI)-powered computer-aided detection (CADe) systems are emerging as a tool to enhance polyp identification during colonoscopy.

Purpose of the Study:

• To review the role of CADe systems in improving ADRs and reducing adenoma miss rates (AMRs).
• To discuss the broader clinical implications and future directions of CADe in colonoscopy.

Main Methods:

• Systematic review of studies evaluating AI-powered CADe systems in colonoscopy.
• Analysis of CADe's performance in detecting various polyp types and its impact on ADRs and AMRs.

Main Results:

• CADe systems consistently improve ADRs and reduce AMRs, especially for diminutive polyps.
• Limited efficacy of current CADe systems in detecting advanced adenomas or sessile serrated lesions.
• Challenges include potential operator deskilling and the need for algorithm refinement.

Conclusions:

• CADe shows potential to standardize polyp detection quality during colonoscopy.
• The long-term impact of CADe on CRC incidence and mortality requires further investigation.
• Future research should focus on enhancing CADe technology and evaluating its effectiveness in reducing the global CRC burden.