Project Lead: Rehg, Jim Primary Investigator: Kumar, Santosh TR&D1: Enabling the Discovery of Temporally-Precise Intervention Targets and Timing Triggers from mHealth Biomarkers via Uncertainty-Aware Modeling of Personalized Risk Dynamics Lead: Dr. Jim Rehg, Georgia Tech; 10% effort (0.9 CM) Abstract: The mHealth Center for Discovery, Optimization & Translation of Temporally-Precise Interventions (the mDOT Center) will enable a new paradigm of temporally-precise medicine to maintain health and manage the growing burden of chronic diseases. The mDOT Center will develop and disseminate the methods, tools, and infrastructure necessary for researchers to pursue the discovery, optimization and translation of temporally- precise mHealth interventions. Such interventions, when dynamically personalized to the moment-to-moment biopsychosocial-environmental context of each individual, will precipitate a much-needed transformation in healthcare by enabling patients to initiate and sustain the healthy lifestyle choices necessary for directly managing, treating, and in some cases even preventing the development of medical conditions. Organized around three Technology Research & Development (TR&D) projects, mDOT represents a unique national resource that will develop multiple technological innovations and support their translation into research and practice by the mHealth community in the form of easily deployable wearables, apps for wearables and smartphones, and a companion mHealth cloud system, all open-source. TR&D1 will develop, evaluate and disseminate methods to analyze population-scale multi-modal time series of mHealth biomarkers to enable research on identifying the momentary risk factors and risk dynamics that drive adverse health outcomes, while accounting for the uncertainty and missingness inherent in these data sources. TR&D1 will do this under three aims.
Aim 1 will address missing sensor data in mHealth field studies and develop state-of-the art imputation models using novel deep probabilistic neural networks that leverage the hierarchical nature of biomarker computation graphs.
Aim 2 will address compressing a collection of biomarkers that serve as risk factors for a particular adverse health event into a single risk score, to support the online adaptation of decision rules in TR&D2, using longitudinal data that include multiple instances of adverse events and their contexts. In addition to risk scoring, we will also develop models for receptivity to intervention and participant engagement, which complement the assessment of risk in guiding intervention design.
Aim 3 will begin to tackle the critical issue of providing model-based tools for identifying which potential risk factors actually impact risk in different contexts for different individuals, in order to support the intervention design process. TR&D1 will work with its collaborative projects to ensure that it focuses on the most pressing problems facing the mobile health research community. TR&D1 will disseminate its technologies to service projects and the community as software packages and cloud-based data analysis tools, to ensure the usability of these technologies by investigators who are external to the mDOT investigating team. TR&D1 will synergistically work in partnership with the other TR&D projects, the Training and Dissemination Core, and the Administration Core to maximize both the research and societal impact of TR&D1 technologies. 1
Project Lead: Rehg, Jim Primary Investigator: Kumar, Santosh TR&D1: Enabling the Discovery of Temporally-Precise Intervention Targets and Timing Triggers from mHealth Biomarkers via Uncertainty-Aware Modeling of Personalized Risk Dynamics Lead: Dr. Jim Rehg, Georgia Tech; 10% effort (0.9 CM) Public Health Relevance: TR&D1 will enable the identification of momentary risk factors and risk dynamics from mHealth biomarkers that drive adverse health outcomes, while accounting for the uncertainty and missingness inherent in real-life mHealth data. TR&D1 will enable the improved understanding of chronic disease progression, and increase the robustness, personalization and temporal precision of mHealth interventions, thereby boosting the potential for mHealth interventions to significantly improve treatment efficacy and enhance the success of long-term health behavior change and maintenance. 2
