A portable surveillance device powered by machine learning, titled FluSense, can detect coughing and crowd size in real time and then analyze the data to directly monitor flu-like illnesses and influenza trends, according to researchers from University of Massachusetts Amherst.

The new edge-computing platform, envisioned for use in hospitals, health care waiting rooms, and larger public spaces, may expand the arsenal of health surveillance tools used to forecast seasonal flu and other viral respiratory outbreaks, such as coronavirus disease 2019 (COVID-19) or severe acute respiratory syndrome.

The inventors of FluSense partnered with George Corey, MD, executive director of University Health Services; biostatistician Nicolas Reich, director of the UMass-based CDC Influenza Forecasting Center of Excellence; and epidemiologist Andrew Lover, a vector-borne disease expert and assistant professor in the School of Public Health and Health Sciences.

FluSense uses a low-cost microphone array and thermal imaging data with a Raspberry Pi and neural computing engine, not storing any personal identifiable information, such as speech data or distinguishing images.

The research team first developed a lab-based cough model, followed by training the deep neural network classifier to draw bounding boxes on thermal images representing people and counting them.

From December 2018 to July 2019, the FluSense platform collected and analyzed more than 350,000 thermal images and 21 million non-speech audio samples from the public waiting areas of 4 different waiting rooms at UMass’s University Health Services clinic.

The researchers found that FluSense was able to accurately predict daily illness rates at the university clinic, with multiple and complementary sets of FluSense signals “strongly correlated” with laboratory-based testing for flu-like illnesses and influenza itself.

According to lead author and PhD student Al Hossain, FluSense is an example of the power of combining artificial intelligence with edge computing, bringing this kind of technology to the frontier. “We are trying to bring machine-learning systems to the edge,” Hossain said in a press release. “All of the processing happens right here. These systems are becoming cheaper and more powerful.”

For next steps, researchers want to test FluSense in other public areas and geographic locations.

“We have the initial validation that the coughing indeed has a correlation with influenza-related illness,” Lover said in a press release. “Now we want to validate it beyond this specific hospital setting and show that we can generalize across locations.”

REFERENCE
Portable AI device turns coughing sounds into health data for flu forecasting [news release]. Amherst, Mass; University of Massachusetts Amherst: March 19, 2020. https://www.umass.edu/newsoffice/article/portable-ai-device-turns-coughing-sounds. Accessed March 24, 2020.