Smart Speaker Tool Developed that Could Save You from Cardiac Arrest

News Smart Speaker Cardiac Arrest Featured

Last year Apple made a big splash when they announced that a new feature of their Apple Watches would be the ability to help predict heart difficulties, such as atrial fibrillation.

Now smart speakers may be getting in that game as well. Researchers have developed a smart speaker tool that could warn others that you could be having cardiac arrest.

Life-Saving Smart Speaker Tool

While smart speakers have been taking a beating lately with regard to their vulnerabilities, University of Washington researchers created an artificial-intelligence tool for smart speakers that could warn emergency services that you are about to have a heart attack.

When a person is going through sudden cardiac arrest, they will begin by struggling with irregular gasps of breath. That is what is known as agonal breathing. Because of this, they may be unable to speak yet may also be alone at a time when every moment counts and when they are in need of life-saving medical help.

This smart tool that was created by the researchers for smart speakers and possibly smartphones as well. It would be listening for those changes in your breathing. Should it recognize something irregular, emergency services or possibly your loved ones would be alerted to send help.

“This kind of breathing happens when a patient experiences really low oxygen levels, explained Dr. Jacob Sunshine, an assistant professor of anaesthesiology and pain medicine.

“It’s sort of a guttural gasping noise, and its uniqueness makes it a good audio biomarker to use to identify if someone is experiencing a cardiac arrest.”

This tool was developed by using the real recordings of agonal breathing that were captured during emergency calls that were received through Seattle’s Emergency Medical Services.

2.5-seconds of audio were clipped from calls that were recorded on different smart devices between 2009 and 2017. The devices included an Amazon smart speaker, as well as an iPhone and Samsung Galaxy.

News Smart Speaker Cardiac Arrest Alexa

To be sure the tool learned to listen for these sounds in normal everyday conditions, these audio clips were played back from different distances and also had background noises, such as a barking dog or honking car, added. Normal sleeping noises, including snoring and sleep apnea, were added as well to help distinguish real cardiac arrest.

Ninety-seven percent of the time the tool was able to detect agonal breathing with the smart speaker or other devices up to six meters away, according to the researchers.

“A lot of people have smart speakers in their homes, and these devices have amazing capabilities that we can take advantage of,” said Shyam Gollakota, co-corresponding author and associate professor at the University of Washington.

“We envision a contactless system that works by continuously and passively monitoring the bedroom for an agonal breathing event and alert anyone nearby to come provide CPR,” he continued. “This could run locally on the processors” that are in an Amazon, Google, or other smart speaker.

“It’s running in real time, so you don’t need to store anything or send anything to the cloud,” he noted.

More work is needed to avoid false alarms to emergency services. “We don’t want to alert either emergency services or loved ones unnecessarily, so it’s important that we reduce our false positive rate,” added the first author, doctoral student Justin Chan.

Does This Cancel Out Security Vulnerabilities?

We’ve been hearing about security vulnerabilities, particularly with Amazon Echos that utilize the Alexa virtual assistant. We know that they’re always listening and absorbing all the information that they hear, which is many times private information that we don’t want saved on servers somewhere.

However, this is a good use for Alexa. It’s good for it to be listening if it could potentially save your life. Yet, it doesn’t negate the times we don’t want Alexa to be listening. Perhaps they can find a way for it to have safer listening to not make you vulnerable yet also potentially save your life.

Would a smart speaker possibly alerting you of a cardiac arrest increase your desire to buy one? Tells us your thoughts and concerns in the comments below.


  1. “Would a smart speaker possibly alerting you of a cardiac arrest increase your desire to buy one?”
    Having gone through two “cardiac events” heart monitoring would be heaven sent. In fact medical monitoring of any kind is the perfect job for smart speakers. In theory and/or in a perfect world. In reality, I would not touch one with the proverbial ten foot pole. Not as long as Amazon, Google, Microsoft, Apple, etc are in control and are recording and storing every sound.

    As I’ve commented before there could be serious legal ramifications to company employees listening. Do they have the legal obligation to notify somebody (and who) when a user is having a medical event? Can they and the company be held legally liable for not acting quickly enough, and what IS “quickly enough”? There have been suits claiming that paramedics did not respond fast enough and the patient died. LEOs have been accused of not arriving fast enough at the scene of a crime. Are Amazon, Google, et al. willing to into all of that or do they feel that they can gain more than they can lose? IOW, make money off of human suffering.

    1. “IOW, make money off of human suffering.”

      Do you think the hospitals and paramedics are acting out of the goodness of their hearts? It’s a job for them just like any other. I think as someone who’s had a bypass recently I am intrigued by the possibilities of arresting cardiac failures using technologies. Let us give credit where it is due.

      Well look at it another way I do have mixed feelings about big tech companies having access to my private data that would arrange such a life saving mission.

      1. “Do you think the hospitals and paramedics are acting out of the goodness of their hearts?”
        Aren’t you the cynic! 🙂
        But, as you say, it is their job, the primary aim of which is to save lives. It is NOT the job of Amazon, Google, etc whose primary, and probably the only, goal is to make profits. Doctors and paramedics I have to trust. I DO NOT have to trust (and I don’t) the tech companies especially when it comes to my health and especially when run by people like Zuckerberg. Not after the genetics companies tried to patent each and every gene in the human genome.

        BTW – the vast majority of the paramedics in my area are volunteers. The volunteer ambulance corps have very few full-time, paid staff.

    2. Upon review, a smart speaker would not have helped me during either of my cardiac events. They both occurred outside of the convenient listening range of a speaker. So do most other heart attacks. Therefore, while a smart speaker might be of help when a heart attack happens to occur within it listening range, to market them as a life-saving device is a little bit disingenuous. What would have helped with my second heart attack would have been a wearable monitor. Considering the circumstances, no magical device would have helped with my first heart attack since there was absolutely no indication that I was at risk. The only way it could have been anticipated was if I happened to be hooked to an EKG machine.

      To reiterate, to advertise and sell smart speakers as life saving devices is nothing more than callous money grab on the part of the companies. It is also a way to harvest from the users their medical data, which the companies has absolutely no right to, in direct contravention of the HIPAA Law.

      1. I can see why you’d be hesitant to trust it after you have had two heart attacks.

        Here’s what a paramedic said about the Apple Watch and how it will and won’t help.

        His opinion is that it will help and save lives, but you are right, not all the time. I don’t think they are marketing any product to say it will always save your life. I think they are marketing them toward they could save your life. With a history of heart trouble on both sides of my family and with a short history of it as well after my cancer, I would be inclined to get a device that would help protect me, even if I know it was not going to predict it 100% of the time.

        1. Instead of relying on commercial devices designed for other uses (smart speakers, smart watches) which, Oh By the Way, can also monitor for some medical events, and which are meant as revenue generating products, I would rather wear and rely on a purpose built medical monitoring device. Besides, the user would have to pay for commercial devices while a dedicated medical monitor most probably would be covered by insurance.

          “I don’t think they are marketing any product to say it will always save your life.”
          I’m sure they’re not specifically marketing their products as “life savers” for a couple of reasons.
          I may be mistaken but, at least in the US, the laws about marketing and selling devices that have a medical use are much more stringent than devices for common, every day use. Also, the manufacturers do not want to expose themselves to liability when it can be shown that users died when wearing those devices.

          “I think they are marketing them toward they could save your life.”
          That is good marketing. It does not claim that it WILL save your life thus avoiding legal problems when it does not but it suggest to the prospective buyer that it might. After all, who doesn’t want a “guardian angel” watching over them? Having a phrase like “could save lives” or something similar in an advertisement may induce some people, who otherwise may not have any intentions, to purchase the device.

          What about the tendency of many people to presume that all similar devices have similar capabilities? How many people will reason “Apple Watch Series 4 has AFib and fall detection, and ECG capabilities but it requires an iPhone. I can save a lot of money if I buy a Fitbit.” and act on that?

          “With a history of heart trouble on both sides of my family and with a short history of it as well after my cancer, I would be inclined to get a device that would help protect me, even if I know it was not going to predict it 100% of the time.”
          First and foremost, you know as well as I do that no device, not even a doctor, can predict a medical event 100% of the time. Both the device and a doctor are basing their predictions on probabilities. However, a dedicated monitor will have a higher accuracy than an Apple Watch, or any other commercial gizmo.

          AFAIAC, wearable medical monitors should be a priority for IoT device makers. I just hope the monitoring is done by companies experienced in and dedicated to medical monitoring. Amazon, Google, Samsung and others who currently listen in and monitor our smart devices should not be involved in any way. However, being a pragmatist, I know that is not about to happen. Smart toasters and smart toilets are much sexier than smart medical monitors. And based on their past actions, the companies I mentioned will find a way to horn in on the action.

          1. I actually wore a monitor for several weeks last fall to see if I had A-fib or not. It saw that I had issues but not related to A-fib. They are going to keep their eye on me. But that monitor was horrendous! Huge and multiple wires going up to sticky things on my chest. A total pain, difficult to wear under clothes. I wished I could have worn an Apple Watch instead that would have given them the same data.

  2. “But that monitor was horrendous!”
    Bummer! With all the advances in technology and microminiaturization one would think somebody would have come up with a small wearable monitor. Maybe your doctor did not have the latest model.

    As I said, wearable medical monitors should be a priority for IoT device makers. I guess it all comes down to revenue and profits. Maybe they don’t see medical devices as money makers. I am surprised that the traditional medical device makers haven’t yet come out with small monitors. It seems that all they have to do is to reprogram a fitness monitor-like device to gather more data. I’m sure many segments of the medical community are clamoring for devices that are comfortable to wear and can be used to monitor patients 24/7/365.

Comments are closed.