Wednesday, February 01, 2017 4:00pm - 5:00pm
Watson CIT - SWIG Boardroom (CIT241)
Dr. Samantha Kleinberg
Stevens Institute of Technology
From EHRs to body-worn sensors: Finding causes of health and disease
Massive amounts of medical data from electronic health records are now being mined by researchers, and can be used to improve care such as by finding factors affecting the recovery of patients in intensive care, or early risk factors for heart failure. However, patients are highly heterogeneous, the data are noisy, and we lack ground truth for evaluating algorithms. Further, many important events occur outside of medical settings. This is particularly true for chronic diseases such as diabetes and obesity. In this talk I discuss why outpatient data are needed to understand chronic disease, the computational challenges of inferring causality in these data and why we need automated monitoring of nutrition. In the second part of this talk I discuss our recent work using multiple sensing modalities to automatically identifying eating and how this can ultimately be used to support individuals with chronic disease.
Hosted by The Ramachandran Group