New Tech Lets You Test Your Blood Sugar Without Needles
Thanks to artificial intelligence, no-prick glucose testing may be the wave of the future for diabetics.
Finger-prick tests are one of the necessary annoyances that come with having diabetes—but a new study says artificial intelligence (AI) may be able to test your blood sugar levels—no needles required.
A non-invasive wearable sensor can detect low glucose levels using what’s called electrocardiogram (ECG), according to a study from researchers at the University of Warwick in England. Basically, the AI in the device can sense electrical signals from your heart and use them to detect when you’re hypoglycemic (meaning your blood sugar is low) without the dreaded finger-prick, per the study.
Currently, there are two ways to check blood glucose. The most common way is with a glucometer, which requires needle pricks for every blood glucose reading. The newer method is a continuous glucose monitor (CGM) which involves a monitor that records blood glucose from interstitial fluid 24/7 over the course of two weeks. However, in this new method, the non-invasive wearable sensors can detect raw ECG signals from a couple of heartbeats and pinpoint when you’re having a hypoglycemic event, according to results from two pilot studies. This works, the study explains, because blood glucose concentration can actually affect your heart’s electrical activity.
"Finger-pricks are never pleasant and in some circumstances are particularly cumbersome. Taking a fingerpick during the night certainly is unpleasant, especially for patients in pediatric age,” says study author Leandro Pecchia, Ph.D., associate professor of biomedical engineering at the University of Warwick. "Our innovation consisted in using artificial intelligence for automatic detecting hypoglycemia via few ECG beats. This is relevant because ECG can be detected in any circumstance, including sleeping."
The study authors plan to do more research on this technology before it’s available publicly for blood sugar monitoring—but when it does become available, the devices may come with another bonus: They may be more cost-effective than typical invasive CGMs.
Blood Sugar Basics
Testing your blood sugar is a crucial part of managing diabetes, according to the Mayo Clinic. That’s because a glucose monitor collects key information about your disease, like how the food you eat and the exercise you do affect your levels, how your medications are working, and more—including whether your levels are too high or too low.
Based on the type of diabetes you have—1 or 2—and other factors, your doctor should give you instructions on how often to check your blood sugar. For Type 1 diabetes, it may be up to 10 times a day. For type 2 diabetes, it may be just twice daily. Even if you’re using a CGM that’s constantly keeping track of your levels, you likely need to do the finger-prick twice a day to make sure your device is working properly.
Until this fancy new technology becomes available, here are some tips from the Mayo Clinic to make sure you’re doing it correctly:
Follow your doctor’s orders and your monitoring device’s manual—each device is different!
Use test strips that are specifically designed for your device and make sure you store them as directed. Don’t use expired ones.
Take a blood sample in the amount directed in your manual.
Make sure to clean your device regularly.
Take your device with you to your appointments with your doc if you have questions—they can help make sure you’re using it properly.
Blood Sugar Testing Information From the Mayo Clinic: Blood sugar testing: Why, when and how. (2018). mayoclinic.org/diseases-conditions/diabetes/in-depth/blood-sugar/art-20046628
News Release on Device from University of Warwick: Artificial Intelligence (AI) can detect low-glucose levels via ECG without fingerprick test. (2020) University of Warwick. warwick.ac.uk/newsandevents/pressreleases/artificial_intelligence_ai
Study on New Glucose Monitoring Device: Scientific Reports. (2020). “Precision Medicine and Artificial Intelligence: A Pilot Study on Deep Learning for Hypoglycemic Events Detection based on ECG.” dx.doi.org/10.1038/s41598-019-56927-5