By now most of us know Alexa can handle a lot of queries, what we are unaware of is the fact that this device could detect when someone is having a heart attack and automatically call for medical help. We know, Alexa isn’t a health tracking device, and we understand that this news sounds a tad weird. But researchers have finally developed an AI tool which alerts Alexa – or any other smart speaker – when someone is having a cardiac arrest.
How Is Breathing Associated With Heart Attacks
At the inception of a cardiac arrest, people often have some telltale symptoms like pain in the arm or chest, heartburn, irregular gasps or shortness of breath. Irregular gasps of breathing are one of the most obvious ones to the outside observer who can hear a gasping sound made by the patient– this is called ‘Agonal Breathing’. This kind of breathing happens when a patient experiences low levels of oxygen, and it sounds like a rough gasping noise.
It is the uniqueness that makes Agonal Breathing a great audio biomarker that could be used to spot when someone is experiencing a cardiac arrest.
Cardiac Arrest Detection Skill – Quite A Discovery
When a person suffers from cardiac arrest, it becomes near to impossible for him/her to call for help. Wouldn’t it be great, if devices that we own could listen to critical indicators and take immediate action? That’s exactly what researchers at the University of Washington did, they developed an early stage artificial intelligence tool which can detect abnormal breathing and dial the emergency number to call for help.
Shyam Gollakota, associate professor at the University of Washington said,
‘A lot of people have smart speakers in their homes, and these devices have amazing capabilities that we can take advantage of.’
Tried And Tested?
Real agonal breathing recordings captured from 911 calls were used to test this tool. Data was collected from 162 calls between 2009 and 2017, and 236 clips were created from the calls. The recordings were captured from Alexa-enabled, iPhone 5s, and Samsung Galaxy S4 devices. After different machine learning techniques were applied, the team came up with 7,316 positive clips.
These sounds are played back with added background noise from diverse distances to ensure this technology can pick out the abnormal breathing noises amongst other sounds. The devices will be deeply trained to distinguish this explicit sound of breath and contact the emergency services to immediately send someone to help.
97 Percent Correct Detection
So far, this technology has been able to detect the agonal breathing correctly 97 percent of the time. However, the researchers say, this algorithm needs a bit more work to prevent false alerts to the emergency services. According to Justin Chan, a Ph.D. student at the University of Washington said,
“We don’t want to alert either emergency services or loved ones unnecessarily, so it’s important that we reduce our false positive rate”
When it comes to monitoring health, smart speakers win by 2 leads over smartphones and wearables; one they are always active and listening and second, they are always connected to power – this indeed makes them more dependable. With this technology in place, a lot of lives could be potentially saved by bringing in the much-needed help right on time.