Research team develops non-invasive glucose monitoring system

A recent study found that a non-invasive glucose monitoring system developed by scientists at Kennesaw State University (KSU) had a 79 percent accuracy rate.

The pilot study, published in the Journal of Medical Internet Research, was led by Maria Valero, an assistant professor of information technology at KSU’s College of Computing and Software Engineering. Having watched her diabetic father prick himself multiple times a day while checking his blood sugar, Valero set out to develop a less invasive and less painful method to monitor glucose levels.

Along with a team of KSU researchers, Valero created a system using a credit card sized computer known as a Raspberry Pi, a portable camera, and a visible light laser. Together, the devices capture high-definition photos of the skin as the light laser passes through the skin tissue. An artificial neural network model then examines the absorption and scattering of light in the skin tissue, and estimates glucose concentration.

For the study, scientists tested the devices on the fingers and ears of eight volunteers. Glucose measurements from the system were then compared with commercially available glucometers.

According to the study’s authors, their results were promising. Images taken from the finger had a 79 percent accuracy rate while images taken from the ear had a 62 percent accuracy rate.

The authors concluded that this pilot study suggests that blood glucose concentration can be estimated with infrared image of human tissue. While more research and development are needed before the device is used commercially, Valero said she hopes this system will allow for less pain and lower risk of infection while testing for blood sugar.

“Our pilot study was very successful,” she said. “We are excited about how this device will help people with diabetes, which affects about one in every 10 people in the United States.”