Creating a Smart Ring to Monitor Health Conditions
ECE Affiliated Faculty Holly Jimison, Assistant Professors Aatmesh Shrivastava and Hui Fang were awarded a $300K NSF grant to create a “Self-powered Smart Ring for Always-On Health Interventions.”
Abstract Source: NSF
Continuous monitoring is becoming increasingly more important in preventive, personalized care supporting just-in-time interventions. New sensor and communications technologies offer opportunities for more proactive models of care that reach people in their homes and everyday lives to improve health behaviors. Although wearable sensors have played an important role in technology-facilitated health interventions, maintaining use for longer than a few weeks has proven difficult for much of the population. One substantial barrier to extended use is the need to recharge the devices, usually on a daily basis for tailored just-in-time messaging and effective interventions. Current techniques for real-time assessment and interpretation of physiological sensors require power hungry data collection, transfer, and interpretation to deliver tailored and timely feedback. Each time a wearable device requires charging, it takes renewed motivation to charge it and additional motivation to put it back on. The goal of this project is to create a multi-sensor comfortable self-powered ring that never needs charging. The team will use stress monitoring and management as an important clinical challenge requiring multiple sensors and just-in-time interactions as an example to test the contributions of a multi-sensor smart ring, as well as intelligent sampling inference and transmission, to provide tailored stress coaching advice without the need for battery charging or device removal. Stress is one of the key threats to the health and productivity of the nation. As a health hazard, it affects all major organs and is associated with many diseases and a reduction in life expectancy. Normally, monitoring stress data and providing real-time intervention would require recharging a wearable device every 6 hours. To address this issue, the research team plans to develop an energy harvesting system on a chip (SoC) with algorithms that minimize power consumption. This system will be embedded within a comfortable waterproof ring that users never have to remove or charge. This novel capability will enable effective health coaching interventions that require continuous engagement and feedback.
In this one-year project, the team plans to develop a prototype self-powered ring with the capability to monitor health and activity variables important in stress management interventions. These variables include heart rate, heart rate variability, and electrodermal activity (an indicator of skin conductance). This research group also plans to measure motion using an accelerometry sensor to provide information in distinguishing physical stress from the target classification of emotional stress. The project will involve two main activities. Firstly, the project will develop algorithms to minimize power consumption to enable perpetually operating sensors for stress monitoring. To accomplish this goal, the project team will first specify the clinical requirements for the monitoring and feedback protocols. This specification will then inform signal sampling and filtering requirements, data fusion specifications, as well as guidelines for approaches to minimizing data storage and transfer. Decision-theoretic algorithms will be used to optimize the sampling and transmission of multiple sensor data from the ring. Existing algorithms will be optimized for stress monitoring and sensor fusion by integrating heart rate, heart rate variability, electrodermal activity and accelerometry to accurately classify stress levels in real time. The second main activity will be to develop an energy harvesting SoC with dynamic performance scaling and fabrication protocols for smart ring system integration and waterproof packaging. The project will develop a high-precision on-chip clock source and advanced power management circuitry to enable implementation of the dynamic decision algorithms and energy utilization harvested from external indoor solar cells. The team will then integrate high efficiency GaAs solar cells, dynamic sampling/transmission SoC, sensor components (photoplethysmography sensors, electrodermal activity electrodes), and wireless modules with advanced flexible encapsulation scheme into an always-on smart health monitoring ring. It is anticipated that the success of this project will generalize to markedly improve health behavior interventions and will transform the next-generation health and medical research through always-connected data, people, and systems.