Researchers Devise A Way For Smart Speakers To Monitor Heart Rhythm

Consumer electronics manufacturers have made a significant push in recent years to make their devices more useful for various health-related issues. One good example is Apple, with its Apple Watch wearables able to record ECGs and other health data. Researchers from the University of Washington have now conducted research that shows smart speakers like the Amazon Echo and Google Home can monitor some healthcare issues from home.Researchers have shown the devices can detect cardiac arrest and monitor the breathing of infants in the past. Scientists have now demonstrated that smart speakers can be tuned to monitor regular and irregular heartbeats without physical contact. The system can send inaudible sounds from the speaker into the room, and based on how sound is reflected to the speaker, identify and monitor individual heartbeats.

The system developed by the researchers uses machine learning to help the speaker locate signals from regular and irregular heartbeats. During testing, the system was used on healthy participants and on cardiac patients in the hospital. In both groups, the smart speaker detected heartbeats that closely matched the beats detected by standard heartbeat monitors. The team says regular heartbeats are easy enough to detect even if the signal a small by looking at a periodic pattern in data.

However, irregular heartbeats are challenging because there is no periodic pattern to search for. However, the algorithms the team developed were able to identify irregular heartbeats in testing. The team is working on the system because heart rhythm disorders are more common than some other well-known conditions. Cardiac arrhythmias are also responsible for some major health problems, including strokes. The highly unpredictable nature of cardiac arrhythmias can be difficult to diagnose.

The system searches were heartbeats when a person is between one and two feet in front of the smart speaker. A continuous inaudible sound bounces off the person and returns to the speaker. The algorithm can determine how the returned sound changes and isolate the person's movements, including chest movement. The team does point out that they do not receive the electrical signal when the heart is contracting with their method. Instead, they see the skin's vibrations when the heart beats.