Research

UNCW Computer Scientist Collaborates with Others to Analyze Fitness Tracker Data to Predict Consumers’ Health and Longevity 

THURSDAY, NOVEMBER 10, 2016

Karl Ricanek

Using sophisticated algorithms developed at the University of North Carolina Wilmington, researchers now have the ability to translate data from wearable fitness trackers into reliable measures of consumers’ health and longevity.

Karl Ricanek, professor of computer science at UNCW, and S. Jay Olshansky, professor of epidemiology and biostatistics at the University of Illinois at Chicago, presented their findings in a new paper published Nov. 1 in Computer, a publication produced by the IEEE Computer Society, an international organization dedicated to computer science and technology.

“What we are doing is groundbreaking,” Ricanek said. “We are figuring out how to translate raw health sensor data into information that is more beneficial to consumers than number of steps or calories burned. We can now translate steps into measures of life expectancy and decreased morbidity. Sleeping patterns, which are also captured by these Internet of Thing (IoT) activity devices, can be examined for early signs of health or mental conditions.”

Wearable fitness trackers, such as Fitbit and Misfit, collect information on the number of steps users take each day. Some models also monitor sleep patterns, blood pressure, heart rate, pulse and other health-related measurements. Among consumers, these trackers collectively gather trillions of data points daily.

Ricanek and Olshansky collaborated on the “Computer” paper with Bruce Carnes, University of Oklahoma Health Sciences Center; Claire Yang, University of North Carolina at Chapel Hill; Hiram Beltran-Sanchez, University of California Los Angeles; and Norvell Miller and Janet Anderson of Lapetus Solutions Inc.

Using data on the number of steps users take daily in conjunction with other significant information such as age, weight, height and gender, the researchers developed projections regarding the impact of exercise on users’ longevity. When activity and health sensor data are added to “selfie” analysis – research known as facial analytics – the potential increases. Facial analytics developed and patented by Ricanek can determine age, sex, rate of aging, body mass index and smoking status. With sensor data and a “selfie,” public health researchers can better understand the health and wellness of the population. The information may be financially valuable to consumers, as well as health insurers, life insurers, mortgage companies and market researchers.

Each of the researchers is affiliated with Lapetus Solutions Inc., a Wilmington-based company created as a spinoff of Ricanek’s academic research at UNCW and Olshansky’s work at UIC.

“It is exciting to think about UNCW being on the frontier of developing algorithms to understand this data and convert it into useful information,” he said.

--Andrea Monroe Weaver