Smart speakers can be used in the field of health. For example, researchers at the University of Washington have demonstrated the possibility of detecting cardiac arrest or monitoring the baby’s breathing through these products, and now, for the first time, they have been able to develop an artificial intelligence system. Develop Which monitors the heart rate through a smart speaker without physical contact.
The new AI system, by researchers at the University of Washington, emits inaudible sounds through a smart speaker in a room, and can detect and monitor people’s heart rate as it returns to the speaker.
Because the heartbeat is a small movement in the chest area, the team used machine learning to help position signals with a smart speaker to be able to detect regular and irregular heartbeats.
The researchers tested the system on healthy people as well as heart patients in hospitals to test its function. More precisely, the prototype of this system was tested on 26 healthy individuals and 24 heart patients in hospitals. This smart speaker was able to detect heart rate almost similar to standard heart rate monitors.
While many people are familiar with the concept of heart rate, doctors are more interested in diagnosing and evaluating heart rhythm. Heart rate is the average heart rate over time, while heart rhythm describes a pattern of heart rate.
For example, a person with a 60 beats per minute has a normal rhythm and his heart beats once per second. But irregular heart rhythms have different heart rates. Arrhythmias or irregular heart rhythms are more common than other heart diseases and can cause diseases such as stroke, but they are unpredictable and cannot be easily diagnosed.
In this regard, the development of a low-cost device that can be used at home to monitor people’s heart rate can have a significant impact on the condition of people in the stages of heart disease.
When the person is within 30 to 60 cm of this smart speaker, the search process for the heartbeat begins. The AI-based system continuously emits inaudible sounds that monitor the heart rate when it returns to the speaker.
This product uses a self-monitoring machine learning algorithm that combines signals received from all microphones to detect heart rate signals. The system then uses a second algorithm to split the signal into heartbeats to monitor them.