Researchers have developed a method to mine mountains of data provided from wearable sensors to monitor stress in real time and determine the best time to intervene with a text message or prompt.
The Center of Excellence for Mobile Sensor Data-to-Knowledge, or MD2K, headquartered at the U of M, with researchers at the National Institute of Drug Abuse Intramural Research Program officially initially reported their research in a paper last month at the 2016 CHI conference for human-computer interaction in San Jose.
The National Institute of Health in 2014 made MD2K, with U o fM computer science professor Santosh Kumar as center director, one of 11 Big Data Centers of Excellence designed to harness exploding amounts of data to improve the nation’s health and medical care. MD2K involves top scientists from a dozen universities to develop tools, software and training materials for researchers, health care providers and others.
The new research, “Finding Significant Stress Episodes in a Discontinuous Time Series of Rapidly Varying Mobile Sensor Data,” tracked body responses, GPS and activity data in a study of 38 opioid-dependent drug users to identify major but usually rare stress episodes. A standard for continuous assessment of stress researchers developed earlier, known as “cStress” also played a role.
The research also pinpointed the environments, such as areas with graffiti, trash, broken windows and bars, where stress spiked and found that one episode increased the likelihood of another. Through early detection, intervention and prevention, the research could help reduce health care costs and combat diseases and behaviors ranging from obesity and smoking to substance abuse, asthma and congestive heart failure.
©2016 The Commercial Appeal (Memphis, Tenn.) Distributed by Tribune Content Agency, LLC.
