News

The team is led by Associate Professor of Computer Science Cindy Marling.

 

"Everybody's blood glucose levels fluctuate throughout the day whether or not they have diabetes.  But patients who have diabetes have greater fluctuation than normal. Diabetes makes your blood glucose levels rise and so you may take insulin to bring them down," said Marling. "The problem is when they come down, they can be too low, and you can be subject to hypoglycemia, which is a dangerous condition and you can get sweaty and nervous and shaky.  If not treated in time, you can actually lose consciousness or have a seizure. It's very important to be able to predict in advance that this problem is going to occur."

 

Marling says the team aims to use artificial intelligence to predict blood glucose levels.

 

"We have patients giving us their life event data via their smart phones on a daily basis so we need to know what's going on in the world around them: what they're eating, whether they're exercising, whether they are awake or asleep.  And then we get blood glucose signals through sensors that are directly attached to the patient. We then match the events to the signals and try to predict what's going to happen next," said Marling. 

 

In order to predict blood glucose levels, the team is building models specific to each patient.

 

The researchers are using something called support vector regression, which is more commonly used to predict stock market prices and utility loads. 

 

The team is the first to use this process for the medical profession with human subjects.

 

Marling explains that being able to predict blood glucose levels means the patient can act before problems occur.

 

She says the team's research goal is to predict blood glucose levels 30 to 60 minutes in advance and give patients a chance to take preventative steps.

 

"We have hopes of actually using this in a medical device, so if a patient were to be going low overnight, while they were asleep, we could alert them, alarm them, wake them up so they could adjust their insulin to avoid the problem of hypoglycemia that I told you about before," said Marling.

 

The research team consists of biomedical engineering researchers from Ohio University's College of Engineering and Technology and College of Osteopathic Medicine.

 

The researchers recently received a $350,000 grant from the National Science Foundation. 

 

Ohio University says the project forms the cornerstone of a new SmartHealth Lab in the Russ College, which will facilitate additional Smart Health and Wellbeing research. 

 

The researchers have been performing extensive research with patients since 2004, thanks to three internally funded studies, gathering data and testing artificial intelligence software for diabetes management.  The new grant will provide

funding for the next three years, but the team's plans run for a longer term.

 

Marling says, "We'll work on it until we retire, and we're not retiring any time soon."