Keep WETM TV

Smart assistants like ‘Alexa’ could learn to detect cardiac arrest

Local News

(NBC) – About a half a million people in America die each year from cardiac arrest. In many cases, it’s because they didn’t get help fast enough. Now there’s a way that smart assistants like “Alexa” could change that.

New research from the University of Washington shows a way to use machine learning to listen for signs of cardiac arrest.

“By taking advantage of a unique sound people make, a unique type of breathing that people make, when they experience a cardiac arrest.”​​​​​​ said Dr. Jacob Sunshine,  M.D., University of Washington

It’s called agonal breathing, and researchers say an algorithm can now identify it with 97-percent accuracy within a range of about 20 feet, which is the typical size of a bedroom. Devices like the Amazon Echo or Google Home could potentially use the app to listen for that dangerous warning sign.

“Our vision is really that these smart speakers can reside passively in the bedroom and continually monitor breathing.” said Justin Chad, a PHD student at the University of Washington.

If you’re concerned that ‘monitoring’ sounds like spying, researchers say the app was built with privacy in mind. “No data is sent to the internet or 3rd party services… all the processing is done locally on the device.” said Justin Chad

Before it calls 911, the device would issue a warning with a chance to cancel before emergency services arrive. 

In cases of cardiac arrest, faster access to CPR could be the key to survival. The new tool can only flag cardiac arrest when agonal breathing is present, which occurs about 50-percent of the time.

There are no details yet of when the technology will be available to consumers.
 

Copyright 2019 Nexstar Broadcasting, Inc. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.