Join the Department of Chemical and Biological Engineering for a talk titled “The Emerging Crystal Ball: AI in Preemptive Medicine” from Assistant Professor Ishanu Chattopadhyay, Ph.D., of the University of Chicago. The seminar will be held from 3:15–4:30 p.m. March 29, 2023, at Perlstein Hall auditorium, Room 131.
Biography: Ishanu Chattopadhyay, Assistant Professor of Medicine at the University of Chicago, is an expert in artificial intelligence, machine learning, and the computational aspects of data science. Leading the laboratory for Zero Knowledge Discovery (zed.uchicago.edu), Chattopadhyay is interested in unraveling etiologies of complex diseases, and understanding rare event dynamics in natural and human systems. His research has been funded by the US Department of Defense (DARPA), the National Institute for Health, the Alzheimer’s Association, and the Neubauer Collegium for Culture and Society, and his work has been published in top peer-reviewed journals including the Proceedings of the National Academy of the Sciences (PNAS), Nature Medicine, Nature Human Behavior, the Journal of the American Heart Association (JAHA), and journals of the American Medical Association (JAMA). Dr. Chattopadhyay won the prestigious Young Faculty Award from the Defense Advanced Research Projects Agency (DARPA) in 2020 for his work on formal methods to study cognitive dissonance and opinion dynamics. Chattopadhyay’s publications are listed at https://zed.uchicago.edu/publications_by_type.html.
Abstract: Rapid progress notwithstanding, AI and its role in decision making has not reached its full potential. Often the issue is not-enough data, not enough data of the right kind, not enough clarity on the implications of decisions guided by AI, biased data that leads to flawed decisions, and uncertainties—exacerbated by the high sample complexity (data hunger) of the existing algorithms, and subtle hidden foundational assumptions in core frameworks that limit our predictive reach. These issues are particularly important in biomedicine, especially in the context of screening and diagnosis of complex diseases. In this seminar we discuss these key hurdles, and present important advances made in the Chattopadhyay lab (the ZeD lab) with respect to universal screening tools for diverse disorders ranging from autism to dementia to pulmonary fibrosis, where AI is beginning to identify new underutilized diagnostic modalities, going beyond known risk factors.