using artificial intelligence for diabetes health coaching
HHS Diabetes Researcher Diana Sherifali (left) is teaming up with HHS CREATE and its founding director, Jeremy Petch (right) to explore whether a type of artificial intelligence called machine learning can provide daily support in health issues for people with type 2 diabetes.
The recipe for a healthy lifestyle is to eat well, stay active, reduce stress and take prescribed medications. For someone with type 2 diabetes, small changes in any of these categories can have big impacts – positive or negative.
Living with diabetes requires ongoing access to diabetes care to help manage the disease, so patients meet with their healthcare team about once every three months. But what if they have problems between dates?
Diana Sherifali, a researcher at Hamilton Health Sciences (HHS), set out to find out if it was possible to prevent a small problem from becoming a bigger problem by using artificial intelligence. She is taking the first steps to develop a health coaching algorithm to help people with diabetes. This algorithm could be added to existing fitness or wellness apps that already track diet and exercise.
Diabetes health coaching
As a clinical nurse specialist at HHS, an associate professor in the School of Nursing at McMaster University, and an associate scientist at the Population Health Research Institute (PHRI), Sherifali explored whether a type of artificial intelligence called machine learning could be the solution. PHRI is a joint institute of HHS and McMaster University.
A computer can possibly learn winning strategies for almost any situation.
“People with diabetes manage the disease on a daily basis. This means that at least 95% of diabetes management happens outside of the healthcare system,” says Sherifali, who is also a certified diabetes educator who understands the benefits of health coaching. “It is not realistic that their medical teams help them on a daily basis. But, if a machine learning algorithm is developed and applied to the existing technology these people are already using, it could provide some of the support needed between appointments.
Type 2 diabetes occurs when your pancreas does not produce enough insulin, which regulates sugar in the body, or when your body does not respond well to insulin. There is no cure, so blood sugar must be managed. This is done through dietary modifications, exercise, maintaining a healthy body weight, blood sugar monitoring, and possible injections of medication or insulin.
Digital health experts
If the health coaching algorithm is applied to a wellness tracker, it can use existing diet and exercise data. Then, once individuals have added their weight, blood sugar, and medications, the algorithm can determine if any changes need to be made and provide recommendations on what to do.
“With my basic knowledge of artificial intelligence, I knew I had to work with CREATE.”
To explore and further develop this idea, Sherifali partnered with digital health and data science experts at the HHS Center for Data Science and Digital Health (CREATE).
“With my basic knowledge of artificial intelligence, I knew I had to work with CREATE, so I approached them with the idea early on,” says Sherifali. “I was excited when I was told it was worth exploring.”
When you play chess on your phone, have you ever wondered how the computer knows how to play? By automatically playing thousands of games and being rewarded for wins and penalized for losses, a computer can potentially learn winning strategies in almost any situation. This approach to machine learning is called reinforcement learning and led to cutting-edge advances in many applications of artificial intelligence, including self-driving cars.
The same approach was used by CREATE for computerized diabetes health coaching, says Jeremy Petch, founding director of CREATE.
“We feed the algorithm health data, recommendations from a medical professional, and outcomes – good and bad,” says Petch. “It allows him to learn the best strategy in all circumstances, just like the computer opponents you play against on your phone.”
In this case, the data includes blood sugar levels, medications, nutrition, physical activity, weight, and stress.
After developing and testing the algorithm, the team determined that it provided accurate initial recommendations.
More data makes a stronger algorithm
“Now that we’ve determined that the algorithm can learn appropriate recommendations for common issues experienced by people with type 2 diabetes, we need more detailed data to continue refining it,” Petch says.
The next stage of the study will teach the algorithm how to provide accurate recommendations with more complex data. Then, it will eventually be tested with clinicians and finally, with patients.
“The challenge is that there is always more data,” says Sherifali. “The algorithm is not intended to replace in-person appointments, so we will need to determine at what stage there is enough data for the algorithm to be effective in coaching individuals through many different issues, while leaving complex scenarios to a medical team. address.”
Since fitness and wellness apps are already well established as a vehicle to house and track the kind of data people with type 2 diabetes are already monitoring, implementing the algorithm in these apps won’t be the hardest part of the project. The first steps – determining if the algorithm will work, are actually the most difficult steps.
So Sherifali says she is delighted that this first step is already successful. This means that health coaching for people with type 2 diabetes could be within reach in the near future.