Apple Watch Can Detect Early Signs of Diabetes with Over 80% Accuracy

Apple Watches can not only keep us more active, but a bit also detects different health problem. So far, they have been able to detect sleep apnea, high blood pressure, and atrial fibrillation, using data from I’s heart rate sensor. Nowadays, the deep learning network of the mobile health data company is adding another disease to the list. It’s one of the most common diseases in the world – diabetes. Although there was a rumor that Google and Apple are working on hardware that could monitor blood sugar levels, Cardiogram’s latest clinical study used only machine learning and the heart rate sensor of Apple Watch to detect if the person has high blood sugar levels. Brandon Ballinger, the co-founder of Cardiogram app, states that the heart rate sensor of this watch can recognize signs of the condition with an accuracy of 85%. We must admit this is pretty impressive for a wrist application. So, here's how Apple Watch can detect the condition.

Apple Watch Can Detect Diabetes - The Study

Researchers from the University of California at San Francisco and Cardiogram made around 200 million heart rate sensor measurements in 14,011 participants to test how accurate and efficient the neural network’s ability is in making a difference between people with and without the condition. The goal of researchers is to help reduce the number of people with undiagnosed condition with this network called DeepHeart.

How It Works

The watch’s heart rate sensor can monitor patterns of beat-to-beat heart rate variability to recognize changes related to the condition. These changes include increased resting heart rates or slower rate after a workout. In fact, the Apple’s Watch heart sensor detects early stages of the condition before recognizing a severe health complication. And, detecting the disease in early stages will enable people to stay with prediabetes instead of diabetes. This is important as prediabetes can be prevented from progressing into the condition with a proper diet and exercise plan. What’s more, there is a behavioral program that can teach people what to eat and how to exercise to keep their blood glucose levels in check. In the next few months, expect app updates with new features that will help people go through the process of detecting if they have a higher risk of the condition. Also, guide toward the right treatment.

The Ultimate Goal

The ultimate goal of the co-founder of Cardiogram is not only to detect health conditions and diseases but to offer intervention and a guide to the right health care. Cardiogram has been actively used by over a quarter-million people, so it was able to apply DeepHeart to identify numerous conditions associated with the heart using consumer-grade hardware. Last May, UCSF and Cardiogram researchers showed that Apple Watch could identify a common heart arrhythmia, atrial fibrillation, which can lead to stroke. Believe it or not, they showed it has an incredible 97% accuracy. After a few months, Apple joined with Stanford for the Heart Health study - a heart arrhythmia detection program. People can download Cardiogram from the iOS App Store for free (it’s 55.2MB).