Joo Heung Yoon, MD is an Assistant Professor of Medicine, of Division of Pulmonary, Allergy, Critical Care, and Sleep Medicine. Joo has been working to develop complex physiologic models that reflects subtle hemodynamic derangement using various machine learning (aka artificial intelligence) techniques. With the NIH T32 then K23 support, Joo developed tachycardia and hypotension prediction models, conducted real-life data collection using a human bleeding model to compute early hemodynamic instability markers, suggested implementation strategies, and designed a novel reinforcement learning structure using implicit Q-learning for shock patients. He is interested in applying his clinical and analytic knowledge base to microgravity environment, where hemodynamic changes upon stress or illness are assumed to be different from the earth. Also, he hopes to deliver feasible prediction algorithms for hemodynamic or acute autonomic issues and user interface for a long-term space flight where not enough resources present, and high-stake decision should be made in a timely manner.
