Wi-Fi Signals May Soon Detect Breathing Abnormalities In People

Your home Wi-Fi router home has one primary purpose, and that is to link you up to the internet. But what if it could detect health conditions as well?
 
Frequencies emitted from the routers bounce off and pass through everything in the environment, including you. And your movements, no matter how small, can affect the signal’s path from the router to your devices.
 
With this, the National Institute of Standards and Technology has devised a way for Wi-Fi to pick up on these minuscule changes in a person’s breathing to determine whether someone is struggling. The study was first published in IEEE Access.
 
The team uses the channel state of information, or CSI, a set of signals sent from a device to an access point, like a router. The CSI signal from the device is always the same, and the access point should know what it “looks” like. However, as these signals move through objects and people, it gets distorted.
 
This distortion is used to determine how the signal changes and, in turn, whether the person in the room is breathing correctly. The Wi-Fi can determine this based on how our body moves as we inhale and exhale.
 
The team tested the hypothesis on a manikin that replicated different respiratory patterns, including normal and healthy, asthma, abnormally slow and rapid breathing, pneumonia, and chronic obstructive pulmonary diseases.
 
The researchers then recorded how the dummy’s chest rose and fell, and how it altered the path of each signal. The data was then sent through deep learning to analyze and differentiate between different breathing problems. 
 
It resulted in an algorithm that the team named ‘BreatheSmart’, which could correctly identify the simulated patterns in the manikin up to 99.54% of the time.
 
The onslaught of the pandemic has brought on a new interest in Wi-Fi being used as a sensing application. Once approved, the program can be used on phone apps and already installed routers so people won’t have to go out and get special ones to track themselves.
 
 
 
 
[via Technology.org and NIST, Photo 181699675 © Oleg Chumakov | Dreamstime.com]