If you wear Apple Watch, AI will "diagnose" hypertension and sleep apnea syndrome: US research results

Apple has many features in the small "Apple Watch" [Japanese version].In addition to the connection to the mobile phone line, there are also the performance that was mounted on GPS, heart rate sensor, and a standard desktop computer for a while.This amazing device does not have a blood pressure gauge or a sleep polygraph test device in the sleep test room.But thanks to machine learning, it may be useful.

One study was announced at a meeting of the US Heart Association (AHA) held in Anaheim, California in November.According to the appropriate machine learning algorithm, Apple Watch's heart rate sensor and pedometer can accurately predict whether high blood pressure or sleep apnea syndrome.Machine learning algorithms can predict high blood pressure with about 80 % accuracy and sleep apnea syndrome with about 90 % accuracy.

Sleep apnea syndrome, where high blood pressure and breathing stop overnight, are common symptoms related to life -death problems such as stroke and heart attacks.However, they often do not notice that they are sick.

Analyze data for more than 6,000 users

It was a joint research team between the University of California and Cardiogram, the University of Calin Francisco and Cardiogram, the University of Calin Francisco, which could predict hypertension and sleep apnea syndrome on Apple Watch.The University of California, San Francisco, provided data from more than 6,000 Apple Watch users who cooperated in healthcare research on mobile.Cardiogram offers an application that analyzes the heart rate data of devices equipped with heart rate sensors, such as Apple Watch.

Cardiograph's engineer team has applied artificial neural networks that companies such as Google use audio to textbooks to interpretation of heartbeats and steps data.Heart rate and steps data, like audio, are a signal that changes over time.The founder of Cardiogram was a former Google employee, who said that he used the experience of working on the voice recognition of Android smartphones and Google Assistant.

This "Deepheart" system is given a series of data on heart rate and steps and information on health status by many people.Cardiogram and the University of California University San Francisco's team published research results in May that DeepHeart could find a way to predict a destructive fibrillation from the Apple Watch data.The results of this study show that if there is one week of data for the mounting, the algorithm can predict high blood pressure with about 80 % accuracy, and a sleep apnea syndrome with about 90 % accuracy.

The new research has increased evidence that the appropriate algorithms could change Apple Watch from a personal trainer to a personal doctor.Apple announced in September that Apple Watch has been conducting research with Stanford University to see if the Apple Watch can detect atrial fibrillation and arrhythmias, which may lead to seizures and heart failure.

Studies that have nothing to do with Apple announced in May have already suggested that detection is possible.In November, medical insurance company Etna announced that it will provide customers with Apple Watch in partnership with Apple with the goal of reducing medical expenses.

Realizes high -precision guesswork without measuring blood pressure and breathing

Doctors will not diagnose hypertension or sleep apnea syndrome (probably not to make such a diagnosis), just looking at the weekly data on smartwatches.The diagnosis of high blood pressure is performed using a blood pressure gauge, and in the case of sleep apnea syndrome, it is necessary to go to a sleep -only clinic or use a home monitoring device.