Masterclass: Advanced GI Endoscopic Imaging (Live/Virtual) | December 2025-VIrtual Masterclass - Endoscopic Imaging - 06 - AI in EGD Quality and Gastric Neoplasia Detection

VIrtual Masterclass - Endoscopic Imaging - 06 - AI in EGD Quality and Gastric Neoplasia Detection - Chiu

Video Transcription

Video Summary

Dr. Philip Chu from the Chinese University of Hong Kong discusses the integration of artificial intelligence (AI) in enhancing esophagogastroduodenoscopy (EGD) quality and early gastric neoplasia detection. Early detection of gastric cancer, predominantly an Asian health issue, allows for less invasive treatment via endoscopic resection. However, traditional EGD has a significant miss rate for upper GI cancers. Dr. Chu emphasizes standardized procedures, comprehensive anatomical examination, risk stratification, and use of image-enhanced endoscopy. AI systems, like the Cerebral AI developed with Endovision, provide real-time feedback to ensure complete stomach coverage and adherence to systematic protocols, improving diagnostic accuracy, especially for trainees. AI-assisted detection systems demonstrate high sensitivity and specificity in recognizing early gastric neoplasia, although most remain in preclinical phases. Challenges include cost, accessibility, lack of guidelines, potential skill degradation among endoscopists, and legal responsibilities. Nonetheless, AI holds promise as a co-pilot to improve diagnostic quality and early cancer detection, with future work needed to optimize cost-effectiveness and clinical integration.

Keywords

Artificial Intelligence

Esophagogastroduodenoscopy

Early Gastric Neoplasia Detection

Endoscopic Resection

Diagnostic Accuracy