Artificial Intelligence Applied to Colonoscopy: Is It Time to Take a Step Forward?

Area of Science:

• Gastroenterology
• Medical Imaging
• Artificial Intelligence

Background:

• Artificial intelligence (AI) is increasingly vital in medicine, particularly in gastroenterology.
• AI computer vision applications are a research priority, including computer-aided polyp detection (CADe) and computer-assisted diagnosis (CADx).

Purpose of the Study:

• To review the current role, applications, drawbacks, and areas for improvement of AI in colonoscopy.
• To explore AI's potential in enhancing colonoscopy quality metrics beyond polyp detection.

Main Methods:

• Review of current literature on AI applications in colonoscopy.
• Analysis of AI's role in assessing colon cleansing, predicting invasion, polyp measurement, and lesion localization.

Main Results:

• AI systems show promise for improving colonoscopy quality metrics.
• Current AI applications include CADe, CADx, and tools for bowel cleansing, polyp measurement, and lesion localization.

Conclusions:

• While AI offers potential benefits in colonoscopy, concerns about cost-effectiveness and a lack of large-scale, multicentric studies with strong clinical outcomes persist.
• Integrating AI tasks into a single quality-improvement device could aid clinical adoption.