Professor Combats Obesity with Better Math

By KATE ANDRIES | Capital News Service

COLLEGE PARK - A Northwestern University professor is taking a stab at making activity recording apps on smartphones more accurate—by employing more sophisticated math.

Apps like Moves or Argus that track exercise and movement are more popular than ever, but don’t capture data all that accurately in certain situations.

Most movement-tracking apps lose a bit of accuracy when the smartphone is carried in a bag or a pocket. Konrad Kording, an associate professor at Northwestern University, and colleagues have developed a more accurate algorithm for activity recording apps.

The algorithm allows apps to predict the location of a smartphone throughout the day—be it on a belt, in a bag, or in your hand—which increases the accuracy of health tracking apps.

“We wanted to see how well activity recognition could tolerate what people did in their everyday life,” Kording said. While the algorithm wouldn’t be integrated in current app technology in the immediate future, it should help improve the accuracy of these health monitoring apps.

Kording’s research is a part of the Center for Behavioral Intervention Technologies (CBITs) at Northwestern University. Opened in 2011, researchers at CBITs work to develop apps for a variety of tech platforms, including smartphones, websites, text messages, even virtual reality.

While a more accurate app may aid in the fight against inactivity and obesity, it is just a small step.

Despite the state’s efforts to curb rising obesity rates, Maryland adults continue to get heavier— 24.9 percent of adults were obese in 2006, while 27.9 percent were obese in 2010. However, Maryland’s obesity rate has yet to hit the nationwide 35.7 percent, according to the Centers for Disease Control.