Refreshments are available starting at 12:15 p.m. The seminar will begin at 12:30 p.m.
Abstract
The computing devices and software tools that mediate our daily life—like smartphones, wearables, and commercial AI assistants—have created a society-scale platform for studying and engaging with entire populations. For example, developing systems to detect, or even predict, symptoms of disease with sensors built into devices that people already use can enable accessible population scale health interventions and insights. In this talk, Joe Breda will discuss three clinical studies of these computing platforms as tools for public health. First, he will cover how biomarkers continuously monitored by smartwatches can be used to measure pre-symptomatic and asymptomatic evidence of immune response to the flu. Second, he will discuss how sensors built into smartphones can be repurposed for accessible fever monitoring. Finally, Breda will discuss how consumer AI assistants can be used to provide accessible differential diagnosis at scale. Together, these systems enable end-to-end diagnostics, from early detection to confirmation to root-cause analysis of symptoms, which can be made accessible to the population through devices and software tools people carry, use, and interact with daily.
Speaker Biography
Joe Breda is a PhD candidate at the University of Washington and a student researcher at Google, where his work focuses on leveraging society-scale computing platforms—e.g., smartphones, wearables, and consumer AI—as extensible public health infrastructure for studying and engaging with entire populations. His research spans technical development of sensing and AI systems alongside the human-factors and translational work needed to align these technologies with the needs of real communities. By applying these platforms to both traditional clinical applications (such as diagnostics and patient monitoring) and broader public health initiatives (like promoting active transport and healthy behavior), Breda’s work imagines a future where the computing we live with does not just make our individual lives more convenient, but can be leveraged across populations to improve collective well-being. His deployment-targeted approach has driven cross-disciplinary health studies both inside and outside the clinic, including: leading a multi-year flu challenge study at the National Institutes of Health using smartwatches to detect pre-symptomatic biomarkers of immune response; partnering with a local cycling advocacy group to crowdsource a passive measure of cyclist safety across the urban road network, work that was featured on NPR’s All Things Considered; and leading the first large-scale evaluation (20,000 users) of a consumer-facing AI for differential diagnosis at Google. Breda’s research has been published in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies and at the ACM Conference on Human Factors in Computing Systems and Conference on Computer-Supported Cooperative Work and Social Computing, and received a Distinguished Paper Award at the 2024 ACM International Joint Conference on Pervasive and Ubiquitous Computing.