CAPS Town Hall and CAPS Methods Core present: Colin Walsh, MD, MA, Assistant Professor of Biomedical Informatics, Medicine & Psychiatry at Vanderbilt Univ.
550 16th St., 3rd Fl., Room 3700
San Francisco, CA 94143
Suicide kills 123 Americans every day and 800,000 people worldwide every year. It is the 10th leading cause of death in the U.S. and the 2nd leading cause of death in those < 34 years old. I will share our experiences incorporating predictive analytics, implementation science, and clinical informatics to catalyze research in mental health. From machine learning to predict suicide to large-scale phenotyping to understand treatment resistance, and from ethical concerns to privacy recommendations, we will describe a research program applying data science with a clinical lens, focusing on some of the most challenging problems in psychology and psychiatry.
Dr. Colin G. Walsh is a practicing internist and clinical informatician who joined Vanderbilt University as Assistant Professor of Biomedical Informatics, Medicine, and Psychiatry in early 2015. His research is focused in predictive analytics applied to vulnerable populations, clinical workflow, and decision support at the point-of-care. His foci of research and operational work are: 1) machine learning/data science applied to use-cases in mental health; 2) utilization optimization and quality improvement; 3) an analytics approach to value-based healthcare. He is Founder and Principal Investigator of the Health Analytics for Risk, Behavioral, and Operations Research (HARBOR) Lab. After undergraduate training in mechanical engineering at Princeton University, Dr. Walsh attended medical school at the University of Chicago. He completed residency and chief residency in internal medicine at Columbia University Medical Center. He studied machine learning and data science in the domain of hospital readmission risk prediction at Columbia University under research mentor, Dr. George Hripcsak. At Vanderbilt, he continues to develop clinically-grounded predictive models using data science approaches on structured and unstructured clinical data. Examples of active projects range from: 1) Machine learning + natural language processing approaches to predict and phenotype risks of suicidality 2) Analytics approaches to support Value-Based Healthcare 3) Visual Analytics + Machine Learning Approaches to predict healthcare utilization to support interventions in Quality and Clinical Improvement, to 4) Algorithms that identify and predict unnecessary healthcare service utilization in Choosing Wisely.