A deep learning algorithm created by NIST can analyze variations in Wi-Fi frequencies to determine whether someone in space is having trouble breathing.
BreatheSmart is an algorithm developed by researchers at the US National Institute of Standards and Technology (NIST), which uses Wi-Fi routers and other devices to ascertain whether a person is having trouble breathing.
Wi-Fi routers transmit radio waves continually for computers, tablets, and smartphones to receive. The signals flow through anything in their path, including people, furniture, and walls, and either bounces off or pass straight through them.
Due to the sensitivity of the frequencies, any motions—including breathing patterns—slightly affect the direction of the signal from the router to a device. BreatheSmart has been taught to recognize and evaluate these changes.
In an effort to aid physicians in fighting the Covid-19 outbreak at a time when patients were segregated from one another and ventilators were in short supply, the algorithm's development team began looking at ways to better track people's health at home.
Prior studies have looked into exploiting Wi-Fi signals to detect movement or humans, but these setups frequently required specialized sensing equipment, and the information from these experiments was typically sparse.
“As everybody’s world was turned upside down, several of us at NIST were thinking about what we could do to help out,” says Jason Coder, who leads NIST’s research in shared spectrum metrology. “We didn’t have time to develop a new device, so how can we use what we already have?”
They developed a novel method for measuring breathing patterns using the already-existing "channel state information," or CSI. A series of signals known as CIS is transmitted from the client (such as a laptop or phone) to the access point (such as the router).
Because these CSI streams are so small—less than a kilobyte—they don't obstruct the channel's ability to carry data. In order to have a thorough understanding of how the signal was changing and be able to analyze the distortion, the team updated the software on the router to ask for these CSI streams more frequently, up to 10 times per second.
The scientists installed a commercial off-the-shelf Wi-Fi router and receiver along with a manikin used to teach medical personnel in an anechoic room to evaluate the algorithm. The purpose of this manikin is to simulate a variety of respiratory disorders, including normal respiration, bradypnea, tachypnea, asthma, pneumonia, and chronic obstructive pulmonary disease, or COPD.
The teams gathered the data from the CIS streams and analyzed it using deep learning.
Deep learning is a subset of artificial intelligence, a sort of machine learning that enhances a computer's capacity to recognize patterns and analyze new data by imitating humans' capacity to learn from their prior behavior.
The researchers developed BreatheSmart using this technique, and it was 99.54 percent accurate in classifying a range of breathing patterns replicated by the manikin.
Most of the work that’s been done before was working with very limited data, said NIST research associate Susanna Mosleh. We were able to collect data with a lot of simulated respiratory scenarios, which contributes to the diversity of the training set that was available to the algorithm.
The researchers expect that in the future, software and app developers will be able to build applications to remotely monitor breathing using the mechanism described in the article as a foundation.
All the ways we’re gathering the data is done on software on the access point (in this case, the router), which could be done by an app on a phone, Coder said. This work tries to lay out how somebody can develop and test their own algorithm. This is a framework to help them get relevant information.
The scientists’ findings were detailed in an article recently published in IEEE Access.
1 Comments
cool
ReplyDelete