Keith Feldman, PhD

Dr. Keith Feldman is an Associate Professor in the Department of Learning Health Sciences at the University of Michigan Medical School. With expertise in artificial intelligence, machine learning and data science, his academic interests lie at the intersection of computational methods and healthcare with an emphasis on neonatal and pediatric care. Dr. Feldman’s research centers on improving the utilization of data generated through clinical care. This includes efforts in translational science and health outcomes, exploring variability between patient conditions, care patterns, and downstream outcomes. A core of such work lies an idea of augmentation, not automation. Rather than applying computational models to replicate clinical decision-making, his work explores how data-driven approaches can enhance it—by identifying what information, if made available, could improve decisions and workflows. He is also engaged in a body of methodological work, developing novel computational approaches to advance the ways in which AI/ML models address complexities from noisy, high-dimensional, and heterogeneous data commonly found in the analysis of real-world healthcare practice.