The mHealth Center for Discovery, Optimization, and Translation of Temporally-Precise Interventions (the mDOT Center) - Primary Investigator: Kumar, Santosh The mDOT Center – Overall Abstract: Providing care for patients with chronic diseases is one of the biggest drivers of the nation’s rising healthcare spending. Many of these diseases are inextricably linked to mutable health behaviors such as excessive stress, lack of physical activity, and addiction. A key strategy for making self-care and preventive health behaviors more achievable has been the integration of passive monitoring into everyday life via mobile sensors and providing personalized information and guidance to patients. But, to have the maximum long-term efficacy, such mHealth interventions must be delivered at the most opportune moments and their content must be dynamically personalized to the individual, their current context, and changing preferences. Our proposed mHealth Center for Discovery, Optimization & Translation of Temporally-Precise Interventions (the mDOT Center) will provide the methods, tools, and infrastructure for researchers to discover, optimize and deploy temporally-precise mHealth interventions. Organized around three Technology Research & Development (TR&D) projects, the mDOT Center represents a unique national resource that will develop multiple methodological and technological innovations, including easily deployable wearables, apps for wearables and smartphones, and software services. TR&D1 will enable the discovery of the timing, targets, and content of novel mHealth interventions by developing generative AI models to continuously estimate the risk of adverse health events, identify key drivers of risk at each moment, and create multimodal trustworthy interventions from large- scale multi-modal time series of mHealth biosignals and biomarkers. TR&D2 will develop autonomous, stable, and data efficient reinforcement learning methods to maximize long-term therapeutic efficacy and engagement for every individual by accelerating the continuous personalization of mHealth interventions using sensor-based biomarkers (including context predictions) in field environments. TR&D3 will develop biosignal foundation models in collaboration with TR&D1, distill them into resource-optimized trustworthy biomarker models to enable large- scale field deployment, and develop digital twin models to enable continuous testing, diagnosis, and optimization of deployed biomarkers and interventions. Together, these mDOT Center deliverables will transform the ability of researchers and healthcare innovators to exploit the exploding range of generative AI and wearables (e.g., AR glasses and smart earbuds) to develop and deliver dynamically personalized and temporally-precise mHealth interventions to individuals. The iterative “push-pull” interaction with collaborative projects (CP) and direct dissemination to service projects (SP’s) will ensure that the technologies developed by our TR&Ds solve real- world problems, are usable by the research community, and have the broadest impact. The mDOT Center’s training and dissemination (T&D) activities will seek to maximize the societal impact of our technologies by promoting the broad distribution and uptake of mDOT Center technologies and resulting software beyond mDOT Center affiliates. Furthermore, a unifying mDOT Center administrative core and optimized operating procedures will facilitate interactions among TR&D researchers and their associated CPs and SPs; enhance the productivity and impact of Center activities; and provide ongoing management, oversight, and planning.
