Benjamin Smarr | The Time of Our Lives: Taking Advantage of Biological Timescales of Change for Digital Phenotyping

ABSTRACT: Biological systems are highly dynamic, with a great deal of variance in most systems occurring in predictable, physiologically-relevant timescales of oscillation: ultradian (~1-3 h), circadian (~24h), ovulatory (days to weeks across mammals), and circannual (yearly). The ubiquity of these cycles is a double-edged sword: disruption of these cycles, as in shift work, light-at-night, or mistimed meals, can cause lasting medical harm and is in fact a major public health challenge at present. However, the presence of these cycles renders novel, high temporal-resolution data flows amenable to data science approaches for discovering oscillation features that enable personal, dynamic digital phenotyping and medical monitoring. In my talk I will describe some of the ways I have been extracting signatures of these cycles from miniature sensors and existing data flows, and using these to detect physiological changes at the scale of populations, but with resolution down to the individual. I will also highlight my work to probe the networks underlying these oscillations, and their potential to enable higher precision for clinical diagnostics, and biomechanistic investigations.BIO: Smarr is an National Institutes of Health K99 fellow in the Helen Wills Neuroscience Institute and Department of Psychology at UC Berkeley, in the lab of Lance Kriegsfeld. Smarr earned his Ph.D. in neurobiology from the University of Washington, and his bachelor of science in biology from UC Santa Cruz . He is the recipient of many awards from the NIH, the National Science Foundation, and other society awards, in addition to building private partnerships with engineering groups and device companies.