Colorectal polyp characterization with standard endoscopy: Will Artificial Intelligence succeed where human eyes...

Area of Science:

• Gastrointestinal Endoscopy
• Artificial Intelligence in Medicine
• Colorectal Cancer Screening

Background:

• The American Society for Gastrointestinal Endoscopy (ASGE) proposed "resect-and-discard" and "diagnose-and-leave" strategies for diminutive colorectal polyps.
• Current community practice often fails to meet the diagnostic thresholds set by ASGE guidelines, leading to suboptimal outcomes.
• Artificial intelligence (AI) offers a potential solution to enhance diagnostic accuracy in endoscopy.

Purpose of the Study:

• To review the recent literature on AI applications for colorectal polyp characterization.
• To evaluate the performance of AI models compared to expert endoscopists and ASGE guidelines.
• To identify current limitations and future directions for AI implementation in clinical practice.

Main Methods:

• Review of recent studies applying deep learning algorithms and AI models in gastrointestinal endoscopy.
• Analysis of AI system accuracy in optical biopsy of colorectal polyps.
• Comparison of AI performance against expert endoscopists and established ASGE diagnostic thresholds.

Main Results:

• AI models, particularly those using deep learning, demonstrate high accuracy in characterizing colorectal polyps.
• AI systems match the diagnostic accuracy of expert endoscopists in optical biopsy.
• AI integration significantly improves endoscopists' diagnostic accuracy and reduces diagnosis time.

Conclusions:

• AI shows significant promise for improving the accuracy and efficiency of colorectal polyp diagnosis.
• AI models currently exceed ASGE recommended thresholds for polyp characterization.
• Addressing current limitations is crucial for the successful clinical implementation of AI in endoscopy.