Masterclass: Advanced GI Endoscopic Imaging (Live/Virtual) | December 2025-Next generation colonoscopy- AI in colorectal neoplasia detection

Video Transcription

Video Summary

This presentation discusses the integration of AI in colonoscopy to improve colorectal neoplasia detection. Colorectal cancer remains a leading cause of death, and traditional colonoscopy misses about 25% of adenomas. AI-based computer-aided detection (CAD) systems significantly enhance adenoma detection rates (by 20-24%) and reduce miss rates by over 50%, especially for small and flat lesions. FDA-approved systems like GI Genius and Fujifilm’s CAT-I provide real-time alerts with minimal additional training required. However, AI use leads to a rise in non-neoplastic polyp detection, increasing unnecessary procedures. Despite enthusiasm, adoption is limited by concerns over false positives, distraction, procedure time, cost, and medico-legal issues. A recent study revealed potential deskilling of endoscopists after AI exposure, highlighting the need for training and competency maintenance. Key knowledge gaps include lack of long-term outcome data, unclear cost-effectiveness, underrepresentation of special populations, and undeveloped legal and ethical frameworks. Addressing these is vital before widespread clinical AI integration to ensure benefits while minimizing risks.

Asset Subtitle

Nasim Parsa

Meta Tag

Concept Artificial Intelligence

Concept Lower Endoscopy

Concept Colorectal Neoplasia

Concept Computer-Aided Detection

Concept Computer-aided Diagnosis

Keywords

AI in colonoscopy

colorectal neoplasia detection

computer-aided detection

adenoma detection rate

medical AI challenges

Artificial Intelligence

Lower Endoscopy

Colorectal Neoplasia

Computer-Aided Detection

Computer-aided Diagnosis