Optimizing Diabetes Care: The Research Supported Promise of AI

The frigid February morning my father left the house to test his blood sugar for prediabetes, he slipped on a patch of ice in the driveway and shattered his leg. 

Though he said nothing at the time, he later told me how wonky his vision had become without him realizing. His first days in the hospital he couldn’t read the writing on the whiteboard 10 feet away. Combining his blurred vision with snow flurries, it’s no wonder he never saw the ice.

It was a painful experience for my dad, but one that forced him and the healthcare system into immediate testing and treatment. Turns out he was even unluckier than we thought – his reversible prediabetes had turned into full blown around-for-the-long-haul Type 2 diabetes.

He’d experienced various symptoms which eventually prompted him to get tested, but like many overworked Canadians, he’d originally chalked up these symptoms to being tired and his challenges with weight loss. He was even drinking more water than ever before and hydration is always a good thing, right? Guess again. Unusual thirst is a symptom of prediabetes and Type 2 diabetes. 

More than 11.7 million Canadians currently have Type 1 or 2 diabetes or prediabetes. “With no signs of [the trend] leveling or decreasing,” Diabetes Canada predicts the ever-rising number will pass 13.5 million by 2030. That’s nearly 35 per cent of Canadians.

Using AI for Diabetes

Research is essential to innovations in the medical field. Insulin, for example, was created thanks to diabetes research. Initially available only in a syringe, now diabetics can opt for an insulin pump instead.

The exciting new development in diabetes research is the application of AI. Though I love ChatGPT for helping me write my article headlines, the world of AI diabetes management is much cooler. 

Researchers have found that AI models have the potential to monitor and adjust your insulin better than your physician can. A study published in Nature found that when the participant’s insulin pump was controlled by AI, they faced fewer severe adverse events like hypoglycemia. 

Not yet approved for use in North America, companies are developing AI devices that learn your eating habits, exercise routine, and blood sugar trends so the technology can adjust your insulin dose accordingly. 

There are diabetes self-management apps in development using AI, too. Most diabetes self-management apps are food focused and allow you to track what you eat. Though helpful, it can be rather time consuming and annoying to input ingredients one by one for accurate results.

Researchers are trying to merge one such app with AI to make diet and lifestyle tracking much easier. Including this tech would allow the app to learn your habits and make recommendations on how to hit your glycemic goals. 

Another app, Nudge, by Rufus Sweeney, MD, uses AI to predict when your glucose is going to spike and gives you recommendations on how to lower it. A continuous glucose monitor is used to measure how your sleep, nutrition, and exercise influence your blood glucose. 

The app is not yet available for download, but it is accepting early applications on their website. 

Taking Care

Between diet, exercise, and lifestyle, managing diabetes is a lot of work. But with the introduction of AI, it doesn’t have to be and it will continue to be easier over time. 

Your doctor cannot know everything happening in your life that impacts your symptoms. AI on the other hand, is showing the potential to follow you throughout life and make recommendations that put you in the driver’s seat of your health.

—————————- Read more from The Health Insider —————————–

~ Read more from The Health Insider ~

Exit mobile version