A new continuous glucose monitor (CGM) is not only non-invasive, but can also predict blood glucose levels from electrocardiogram data, potentially revolutionizing diabetes management. The device will be displayed at the 2024 Taiwan Innovation and Technology Expo. Continuous glucose monitors (CGM) have revolutionized diabetes management by providing 24/7 blood glucose levels and trends and providing comprehensive data that allows medical professionals to personalize insulin therapy.
Essentially, CGM enables better blood sugar control, thereby reducing the risk of disease-related complications in people with diabetes. However, current CGM has some limitations. They are invasive and require inserting a sensor, albeit a very, very thin one, under the skin. Additionally, they are expensive, can only be used for a limited number of days, and generate medical waste. However, at the 2024 Taiwan Innovation and Technology Exhibition, SingularWings Medical Company demonstrated a non-invasive CGM that uses electrocardiogram (or ECG) signals to monitor blood sugar. That's right, it's the heartbeat, the signal from the heart.
"We developed this sensor to collect physiological data, including electrocardiogram, heart rate, respiration and body temperature," Jenny Tseng, marketing specialist at SingularWings, said in an interview with "New Map" at the expo. "At this show, we demonstrated blood glucose monitoring. So, we used electrocardiography to estimate high, medium and low blood sugar levels."
Your skepticism is understandable. How could heart rhythm and blood sugar be linked? How can a device provide continuous blood glucose monitoring without measuring body fluids? How accurate is it? These questions all make sense.
Let's start with the basics: what is an electrocardiogram? It is a non-invasive way of reading the heart's electrical activity, often used to diagnose heart rate or heart rhythm problems. In a clinical setting, it may be necessary to place small gel pads on the chest, arms, and legs and then lie still for a minute or so. Of course, many of today's wearable devices continuously monitor the user's heart rate and rhythm.
Although ECGs are primarily used for diagnostics in the field of cardiology, over the past few years, with the development of wearable device technology such as smart watches and smart rings, researchers have begun to investigate whether ECG signals can be used to estimate blood sugar levels in healthy and diabetic people.
Research shows that low blood sugar (hypoglycemia) and high blood sugar (hyperglycemia) affect the electrical properties of the heart, producing specific changes that can be seen on an electrocardiogram. Researchers use machine learning to analyze electrocardiograms and develop algorithms based on this, using these algorithms to accurately predict blood sugar.
The same goes for SingularWings' non-invasive CGM, although that product is still in development.
"Because we're using a machine learning approach, and ... I can tell you that the accuracy, the average accuracy, is about 80 percent," said Dr. Dick Hsieh, account manager at SingularWings. "But this is related to model training. We still need [a] validation phase. But the results of the validation phase are not known yet."
The user's electrocardiogram is recorded via SingularWings' BEATINFO sensor, which is integrated with the BEATINFO Health app on the Apple Store and Google Play. The sensor weighs only 15 grams (0.5 ounces) and can be worn via a chest strap or a removable patch; both are biocompatible and hypoallergenic.
The blood glucose monitor transmits data to the Health app in real time, which notifies the user if their blood sugar falls outside the normal range. SingularWings expects its non-invasive CGM to be commercially available within two years.