Medtronic, IBM study: Machine learning, app, glucose monitoring hardware combo improves diabetes management

The study​ was based on data from 3,100 people with diabetes who used Medtronic's Guardian Connect glucose monitoring hardware for at least five days.

Science-based healthcare: How IoT and AI can help us make health decisions Philippe Kahn, CEO of Fullpower Technologies and creator of the first phone camera, talks about how the combination of IoT and sensor technology with machine learning and AI is leading to a digital transformation of healthcare and wellness research. Read more: https://zd.net/2GCHRm5

Medtronic and IBM have data showing that the combination of a continuous glucose monitoring system and Suger.IQ app enables diabetics to stay their time in optimal glycemic range for an extra hour a day.

Special Feature

The NHS and technology: How innovation is revolutionizing healthcare (free PDF)

AI and robots, IoT, virtual and augmented reality, and wearables—all are innovative technologies that could boost healthcare and productivity across the NHS. This ebook looks at how these technologies are being implemented and their current and future impact on health services.

Read More

The two companies presented their data at the American Diabetes Association Annual Meeting on Monday. The study was based on data from 3,100 people with diabetes who used Medtronic's Guardian Connect glucose monitoring hardware for at least five days.

Sugar.IQ rolled out in January on Apple's iOS and the app leverages machine learning data from Guardian Connect to provide individual tips for staying in range.

Primers: What is AI? | What is machine learning? | What is deep learning? | What is artificial general intelligence?    

Individuals using the Sugar.IQ app with Guardian Connect along with IBM Watson Health had 4.1% more Time in Range (63.4%) compared to those with Guardian Connect alone (59.3%). Time in Range is the percentage of time diabetics say in the optimal glycemic range of 70-180 mg/dL.

In addition, the study found that Sugar.IQ users found insights helpful 91% of the time.

IBM Watson Health also outlined its prediction model, which was trained on anonymized Guardian Connect user data. The machine learning models had 90% accuracy in predicting hypoglycemic events in a 2-hour window.