Machine learning identifies diabetics with 85% accuracy using wearable heart rate monitors

Machine learning identifies diabetics with 85% accuracy using wearable heart rate monitors

Cardiogram offers an app for both Apple Watch as well as Android Wear and uses a DeepHeart neural network to determine if a person has issues with the heart. Let's wait and see what Android Wear does with this information.

In this study, UCSF and Cardiogram researchers are looking to lay the groundwork for future mHealth wearables with more sophisticated sensors.

"Typical deep learning algorithms are data-hungry, requiring millions of labeled examples, but in medicine, each label represents a human life at risk-for example, a person who recently suffered a heart attack or experienced an abnormal heart rhythm", Hsieh explained in a Q&A. "The final deep neural network contained 564,227 neural network weights and both convolutional and recurrent layers". The researchers were reportedly 85 percent successful in identifying people with prediabetes.

However, Apple Watch has evolved over the years to incorporate advanced heart rate sensors that can detect atrial fibrillation, hypertension, or sleep apnea.

The research tapped into data from 14,000 Apple Watch users and was able to detect that 462 of them had diabetes, with a little help from an artificial intelligence algorithm.

Twitter Posts Its First Profit
But for every silver lining, there's a grey cloud, and Twitter can't rely upon rosy user numbers to bolster the good mood. Shares in Twitter jumped 26% in pre-market trading on Thursday after the social network reported its first ever profits.

According to one of Cardiogram's co-founders Johnson Hsieh, "Your heart is connected with your pancreas via the autonomic nervous system". "As people develop the early stages of diabetes, their pattern of heart rate variability shifts".

mHealth and telehealth advocates have long sought to use technology to identify those people and put them on a healthier path before they either develop diabetes or experience adverse health outcomes because of poor health management.

Apart from Cardiogram, Apple too is engaged in a research of its own - Apple Heart Study - in collaboration with Stanford University to ensure the Apple Watch is able to accurately predict common heart ailments. They were also given a mobile app, which integrated with HealthKit and continuously stored and processed the participants' heart rate steps and daily activity, according to the study. While Apple and Google have been rumored to be working on hardware capable of monitoring glucose levels, Cardiogram's study used nothing more than machine learning and the Apple Watch's heart rate sensor to detect whether a user has diabetes. A 2015 Framingham Heart Study determined that "low" heart rate variability and "high" resting heart rate correlated with the development of Type 2 diabetes over a 12-year period.

Cardiogram's work could eventually prove useful in acting as a pre-screen for these kind of conditions.

The kind of diabetic pre-screening DeepHeart makes possible will eventually wind up in Cardiogram's app, though Ballinger wouldn't confirm when that would actually happen.