ASGE Annual Postgraduate Course: Clinical Challenges and Real-Time Decision Making in GI Endoscopy (Live and Virtual)

Video Transcription

Video Summary

In this video, Yuichi, a Japanese gastroenterologist working on AI research in endoscopy, discusses the concept of computer vision in endoscopy. He explains that computer vision involves using machine learning and AI to optimize the analysis of images and videos in a way that mimics human vision. He highlights various examples of computer vision in endoscopy, including edge detection, optical flow, image stitching, and image processing. He also discusses the role of AI in endoscopy, specifically computer-aided detection, computer-aided diagnosis, and computer-aided quality assurance. <br /><br />Yuichi emphasizes the importance of understanding AI basics and the role of deep learning in endoscopy. He explains that deep learning automatically learns the best features from images and outperforms conventional machine learning methods, especially when there is a large dataset available. He discusses the importance of collecting and separating data into training, validation, and testing sets, and the need to avoid patient-level overlap, data leakage, and exclusion of real-world data during the machine learning process. He concludes by recommending a textbook for GI endoscopists interested in learning more about AI in endoscopy.

Asset Subtitle

Yuichi Mori, MD, PhD, FASGE

Keywords

computer vision

endoscopy

machine learning

AI research

deep learning