Roche introduces AI-powered diabetes tracker to predict blood sugar highs and lows

Roche raised the curtain on its upcoming continuous glucose monitoring system featuring artificial intelligence-powered forecasts of blood sugar highs and lows. 

At the International Conference on Advanced Technologies and Treatments for Diabetes, being held this week in Florence, Italy, the company put forward some initial accuracy data for its new Accu-Chek SmartGuide sensor and apps. 

The real-time CGM is designed to offer AI-based glucose predictions from 30 minutes out to two hours in the future—as well as inform the user of their risk of developing hypoglycemia overnight. Roche is currently working to obtain a CE mark approval in Europe, ahead of a planned commercial rollout in select countries.

“Providing users with predictive capabilities will empower them to proactively adapt their therapy so they can maintain optimal glycemic control and prevent dangerous short- and long-term complications,” Marcel Gmuender, Roche’s global head of diabetes care, said in a statement.

“We know from recent research that glucose predictions provide people with diabetes with invaluable insights to navigate the complexities of the disease,” Gmuender said. “This is one element that can help to lower the burden of daily diabetes management.”

At the ATTD conference, the 14-day Accu-Chek SmartGuide system—designed for people with either Type 1 or Type 2 diabetes, who are taking changing insulin doses over the course of a day—demonstrated an overall mean absolute relative difference of 9.2% in a study of 48 participants. Also known as MARD, the measure estimates the accuracy gap between a CGM’s readings and a reference glucose tracking device, such as a fingerstick reader.

The system includes the wearable sensor and two separate smartphone apps—SmartGuide and SmartGuide Predict, for both Android and iOS—which can display new real-time readings every five minutes, illustrate trends and ring customizable alarms. The predictive features also aim to highlight patterns in glucose readings for additional review by users and caregivers.