March 22, 2020
One of the obstacles to accurately estimating the prevalence of sickness in the general population is that most of our data comes from hospitals, not the 99.9 percent of the world that isn’t hospitals. FluSense is an autonomous, privacy-respecting system that counts the people and coughs in public spaces to keep health authorities informed.This new AI-powered system monitors coughing sounds to understand where the coronavirus is spreading.
The FluSense device first detects coughing and crowd sizes in real-time. It then analyzes the data to predict the progress of COVID-19 and other respiratory diseases.
These insights could guide public health responses to the pandemic, such as the allocation of medical supplies, travel restriction, and vaccine campaigns.
“We believe that FluSense has the potential to expand the arsenal of health surveillance tools used to forecast seasonal flu and other viral respiratory outbreaks, such as the COVID-19 pandemic or SARS,” Rahman told TechCrunch. “By understanding the ebb and flow of the symptoms dynamics across different locations, we can have a better understanding of the severity of a novel infectious disease and that way we can enforce targeted public health intervention such as social distancing or vaccination.”
“I’ve been interested in non-speech body sounds for a long time,” said researcher Tauhidur Rahman, an assistant professor of computer and information sciences at the University of Massachusetts Amherst.
I thought if we could capture coughing or sneezing sounds from public spaces where a lot of people naturally congregate, we could utilize this information as a new source of data for predicting epidemiologic trends.
How FluSense works
FluSense basically consists of a thermal camera, a microphone, and a compact computing system loaded with a machine learning model trained to detect people and the sounds of coughing.
The FluSense system captures cough sounds through a collection of microphones and people via a thermal camera that detects body heat.
The data they collect is processed through a Raspberry Pi computer, which is connected to a neural network that recognizes the sound of coughs. The system then counts the number of coughs and people, to predict where the coronavirus is spreading.
To test the system, researchers have encased the FluSense devices in boxes (the size of a large dictionary) and put them in four healthcare waiting rooms at the university’s hospital.
To be clear at the outset, this isn’t recording or recognizing individual faces; Like a camera doing face detection in order to set focus, this system only sees that a face and body exists and uses that to create a count of people in view. The number of coughs detected is compared to the number of people, and a few other metrics like sneezes and amount of speech, to produce a sort of sickness index — think of it as coughs per person per minute.
Over seven months, it analyzed more than 350,000 thermal images and 21 million audio samples from the public waiting areas.
The system accurately identified coughs 81% of the time. Not bad, but more work will be needed to show, So it can work across different locations.