The Marriage of AI and Diabetes Research Leads to Fascinating Developments

Researchers and other health professionals are implementing artificial intelligence (AI) into their efforts to advance diabetes education and medicine. IBM Watson Health teamed up with Medtronic to develop the Guardian Connect system, which is the world’s first Smart continuous glucose monitoring (CGM)* designed for people who inject insulin daily. It helps people living with diabetes by alerting them that they are about to hit high or low blood sugar levels with over 98% accuracy ten to sixty minutes prior to the actual spike or drop by tracking glucose levels in real time.

Medtronic pairs the Guardian Connect glucose monitor with IBM Watson’s Sugar.IQ app, which provides insights regarding the patient’s typical insulin level trends and patterns so that they can make modifications to their daily regimen and gain more control over the influence that diabetes has on their life. Sugar.IQ uses an algorithm that considers the individual patient’s repeated behaviors and user preferences to solve problems regarding that patient’s particular daily struggles, teach them changes in behavior that could lead to lifestyle improvements, and encourage them to take the necessary steps toward better self-care.

“From this volume of data, it becomes much easier for patients to understand the direct impact that meal choices and carb counting have on their insulin requirements. This will be true for patients with type 1 diabetes (T1D) and type 2 diabetes (T2D),” Dr. Lisa Latts, MD, MSPH, MBA, deputy chief health officer at IBM Watson Health explained to EndocrineWeb. Latts found that a certain patient she was working with was able to physically see how stress affects her blood sugar levels by comparing Sugar.IQ insights found on weekdays to those found on relaxing weekends.

According to Medtronic, those who use Sugar.IQ typically function with a healthy glucose level for thirty-six extra minutes per day than they originally experienced. This improvement, which may appear minor, leads to an average of nine more days a year in which the patient maintains healthy levels.

AI technologies don’t just take general information and feed it into a finite, fixed algorithm. Instead, they employ a process called “machine learning,” in which it uses data to locate trends and patterns, form predictions, and adjust their processes accordingly.

The usefulness of machine learning’s adaptability and pattern recognition is evident in the development of an artificial pancreas, which is perhaps the most exciting possibility regarding the marriage of AI and diabetes research. Instead of trying to create a bionic pancreas that mimics how the average human body reacts to changes in glucose and insulin levels, an AI pancreas will invoke machine learning to understand patterns in that individual patient’s body and make accurate predictions as to how to treat highs and lows accordingly.

The prospects of machine learning provide the possibility of a closed-loop pancreatic system, in which the artificial pancreas is 100% controlled by advanced software that is reacting to a glucose monitor, allowing it to treat unhealthy glucose levels without the patient having to lift a finger. In order to do this, the system will need to be fed a significant amount of information regarding that patient’s glucose and insulin levels, which they can get from CGMs.

“The idea is to train artificial intelligence algorithms on vast amounts of data from diabetic patients, and to use the resulting trained algorithms to run a closed-loop artificial pancreas.” wrote Slate. The technology would ideally be able to monitor and update the algorithm with new trends and patterns as the patient uses the artificial pancreas.

Artificial Intelligence, however, is a growing, developing technology that researchers still don’t have a complete hold of. While AI does provide interesting possibilities int he world of medical research, it still has a little ways to go before we can trust it to take on the full responsibility of the complex functions of an organ like the pancreas. Still, developments like Guardian Connect and Sugar.IQ allow us to remain optimistic about the future of diabetes treatment.