Gastroenterology and Artificial Intelligence: 2nd Annual Artificial Intelligence Summit | October 2020-The Algorithms: An Alphabetical Soup

Video Transcription

Video Summary

In this video, Professor Ulis Bagsy, a faculty member at the Center for Research in Computer Vision and SAIC Chair Professor at the Computer Science Department in the University of Central Florida, discusses the evolution and drawbacks of AI algorithms in medical applications. He emphasizes the need for explainable AI to establish trust between users, physicians, and machine learning algorithms. He showcases examples where AI systems fail, including biases in algorithms and errors in medical diagnosis. To improve the trust and robustness of AI algorithms, he proposes involving physicians in the development of deep learning algorithms through eye tracking technology. Using real-time gaze data, the physician can interact with deep learning algorithms, providing explanations and improving the algorithm's performance. Professor Bagsy also mentions the importance of data quality, removing biases, increasing interpretability, and developing more robust algorithms in the medical AI field. He concludes by highlighting the need to build transparency, trust, and fairness in the application of medical AI. No external credits were mentioned in the video.

Asset Subtitle

Ulas Bagci, PhD

Keywords

AI algorithms

medical applications

explainable AI

biases in algorithms

errors in medical diagnosis