Alexa can determine cardiac arrest by listening to a person’s breathing.

If you are unlucky and your heart stops,the odds will be slim. Of the 400,000 people who encounter this each year in a particular country, less than 6% survive. This is the main cause of natural death. However, the probability of survival will be much higher if you provide immediate assistance: immediate resuscitation can double or triple the chances. A new instrument that uses the microphone of a smart speaker or smartphone to detect timely signals can improve survival rates.

How to save a person with heart failure?

The system developed by scientists from the UniversityWashington uses machine learning to identify what is known as gasping breathing (also known as agonal breathing), which is manifested in a person who lacks air. This is an early warning sign that appears in half of all cases of cardiac arrest.

Scientists have trained the system using recordsagonal breathing received during emergency calls in King County, Washington. They used 729 calls, giving a total of 82 hours of recordings. Then they trained the system on other sounds that can also be heard in the room, like snoring or sleepy muttering, in order to weed out any false alarms.

Two different record sets were used: sleep sounds collected by 35 volunteers and the sounds of 12 patients who participated in the sleep study, because they suffered from snoring and apnea. Recent recordings produced sounds similar to agonal breathing, helping to improve instrument accuracy.

"When we tested it on our system, wefound that the level of false positives in the group of volunteers was 0.2%, and in the sleep study it was 0.1%, ”says Justin Chan, who led the study.

The system was able to correctly determine agonal respiration in 97% of cases at a distance of up to 6 meters.

While the tool is being testedconcept, so there will be many more years before its release to the general public, although scientists plan to eventually commercialize it. Much remains to be verified. Most likely, the system will call an ambulance automatically, but will warn the person about it so that he can cancel the call.

What do you think, how many lives will such a system save? Tell us in our chat in Telegram.