IBM Watson Health, Medtronic app predicts hypoglycemic events with up to 90% accuracy

The artificial intelligence-powered Sugar.IQ diabetes assistant app can help patients stay within optimal glycemic range for longer periods of time and significantly reduce hypoglycemic events, according to study data presented by IBM Watson Health and Medtronic at the American Diabetes Association's Scientific Sessions on June 10.

The app combines IBM Watson's predictive AI capabilities with Medtronic's Guardian Connect continuous glucose monitoring system. The system is thus able to analyze how food, insulin and daily routines impact an individual's glucose levels and predict potential hypoglycemic events several hours in advance.

In the study, more than 3,100 diabetes patients used Sugar.IQ for at least five days. During just that time, on average, users experienced a significant increase in the amount of time spent in their ideal glucose level range — an average of one extra hour per day. The app's predictive alerts also reduced low and high sensor glucose excursions by twofold and fourfold, respectively.

Additionally, per the results of extensive historical data analysis, also presented by IBM Watson Health at this week's event, Sugar.IQ was able to predict hypoglycemic events two hours in advance with 90 percent accuracy and those within a four-hour window with 85 percent accuracy, allowing users to take action well before their glucose levels are out of target range.

More articles about AI:
AI analysis of patient portraits helps diagnose rare diseases
Microsoft building AI hub in Louisville to boost local healthcare industry
Deep learning tool helps radiologists spot brain aneurysms

© Copyright ASC COMMUNICATIONS 2020. Interested in LINKING to or REPRINTING this content? View our policies by clicking here.


Top 40 Articles from the Past 6 Months