Moving Toward Diabetes Solutions That Think for Themselves

One of the exiting things about the race for the artificial pancreas (AP) is that it isn’t on one path—experts in academia and the commercial sector are pursuing many different solutions at once, which advocates for people with diabetes say is a good thing.

Any solution will require data-driven technology to “think” in place of a pancreas that fails to do its job of delivering the right amount of insulin at the right time. But what will drive that technology? More than 40 studies had contributed to the fine-tuning of algorithms over the past decade by 2014, according a review reported in Diabetes Care, and more have occurred since.

A company called Aspire Ventures, which specializes in predictive technology, is taking a different approach. It has created an adaptive artificial intelligence (AI) platform known as A²I, and through its subsidiary Tempo Health, developed a noninvasive system—dubbed Rhythm—to customize glucose monitoring.

Rhythm takes multiple algorithms and creates one that fits the individual patient. And a small study that Tempo Health recently presented shows that the system does a better job than one of the best clinical settings in the world, Diabeter in the Netherlands, which Medtronic purchased in 2015.

Michael Monteiro, MBA, chief product officer at Aspire Ventures, said Diabeter’s “control tower” approach works well at keeping tabs on patients’ blood glucose, but there’s a problem.

“The downside is, it doesn’t scale at all,” he said in an interview with The American Journal of Managed Care. “You have to use AI to sort through the data for each individual, to know what do for that individual at that time of day.”

The study with Diabeter showed that for 7 of 8 participants, the A²I technology resulted in a 20% increase in time in range and a 9% reduction in hypoglycemia,¹ compared with standard care at Diabeter, which is superior to usual care for most people with diabetes.

Tempo Health plans more studies, both for its artificial intelligence platform and for a microfluidic biosensor, which would measure blood glucose through sweat in submicron concentrations, thus eliminating the need for needles. Monteiro said the company sees a future when the person with diabetes will wear a tiny sensor that will capture blood glucose data from a person’s wrist.

“Right now there’s a continuous glucose monitor, an insulin pump, fingers sticks, and an app,” he said. “We want to reduce the form factor to just an Apple Watch and an insulin pump.” This was the vision that DiabetesMine’s Amy Tenderich called for in her “Open Letter to Steve Jobs,” published 10 years ago in April.

Better yet, Monteiro said, there will be fewer delays in measuring blood glucose this way.

A prototype should be available by mid-2017, to be embedded into the clinical trial of the Tempo Health Rhythm system. According to an e-mail from the company, a recent technical breakthrough will allow for reliable and an inexpensive production of the microfluidic biosensors.

Reference

1. Mu D, Funke P, Monteiro M, et al. Biometric data from health patch for glycemic modeling. Presented at Advanced Technologies and Treatments for Diabetes 2017, Paris, France, February 15-18, 2017.