How Artificial Intelligence Is Fighting Diabetes

By Henry Silva

Artificial intelligence algorithms are becoming more and more popular in the field of biology and medicine. A big part of machine learning research these days is devoted to understanding and improving the way we deal with health issues and diseases.

One such issue is diabetes. Today, more than 450 million adults have diabetes worldwide, a number that, according to current trends, will continue to rise. African Americans are more than twice as likely as white Americans to die from diabetes; Latinos in the U.S. have similarly grim outlooks, too. Society has a fixation on high-content sugar products — one that’s going to be tough to break. A worldwide addiction, coupled with other physiological factors has resulted in a vast increase in type 1 and type 2 diabetes.

The good news is that AI is currently getting better at preventing and treating patients diagnosed with both types of diabetes. New research by scientists at POSTECH in Korea found that an AI algorithm could regulate and inject the correct amounts of insulin on diabetic patients way better than the devices currently available on the market. An algorithm like this is a tremendous step towards creating a fully automated artificial pancreas — one that could handle effectively and safely address the physiological needs of diabetic patients.

But are these new AI-based methods an affordable alternative to the methods currently used to treat diabetes?

In the U.S. a large percentage of diabetes patients can’t even afford their insulin, and so the more expensive tech-based methods will continue to remain unaffordable for most Americans. However, there is a growing effort among diabetes patients and their relatives to cheapen the costs of treatment options.

Projects such as Nighscout are trying to create open source alternatives for managing and monitoring data from patients with diabetes, and other big companies are joining in to facilitate patients' access to quality treating methods. IBM’s Watson — an AI created with the objective of understanding and developing new ways of providing healthcare to patients worldwide — is trying to provide more affordable and efficient ways of diagnosing and treating diseases by developing an online environment where patients can be diagnosed and given medical instructions.

Another problem associated with diabetes is diabetic retinopathy — a complication caused by high blood sugar levels that damages the patient’s retina. Leben Care Technologies is using AI algorithms to quickly diagnose diabetic retinopathy, a process that could significantly lower the cost of diagnosing complications associated with diabetes.

If this trend persists, we could have inexpensive and accessible AI-based treatment for conditions like diabetes in the near future. There is no shortage of dedicated scientists working towards more attainable healthcare solutions.