The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus that causes the coronavirus disease (COVID-19), continues to spread across the globe. With over 46.9 million people infected so far, it is crucial to determine how the virus spreads to mitigate its effect.
Many people infected with the SARS-CoV-2 virus are asymptomatic (have no symptoms) or are pre-symptomatic (yet to present symptoms), which means that they can transmit the virus even if they do not feel any symptom at all. With a large fraction of people having no symptoms, it is hard to pinpoint those infected.
Now, a team of researchers at the Massachusetts Institute of Technology (MIT) has found that asymptomatic people may differ from those who are healthy in the way they cough. The differences are inaudible to the human ear but can be detected by artificial intelligence (AI).
In the study, which appeared in the IEEE Open Journal of Engineering in Medicine and Biology, the team reports an AI model that can distinguish asymptomatic people from healthy individuals through forced-cough recordings.
The team believed that COVID-19 patients, especially those who are asymptomatic, could be accurately differentiated only from a forced-cough cell phone recording using AI. They trained the MIT Open Voice model by building a data collection pipeline of COVID-19 cough recordings through their website from April to May. From there, they have built the largest audio COVID-19 cough balanced dataset reported, with more than 5,300 participants.
The AI speech processing framework controls the acoustic biomarker feature extractors to assess COVID-19 from cough recordings. It provides an individualized patient saliency map to monitor patients in real-time. Plus, it is non-invasive and cost-effective.
What the study found
The researchers trained the novel model with thousands of samples of coughs as well as spoken words. When they provided the model with new cough recordings, it accurately identified 98.5 percent of coughs from people with COVID-19, including 100 percent of those from asymptomatic patients.
“The effective implementation of this group diagnostic tool could diminish the spread of the pandemic if everyone uses it before going to a classroom, a factory, or a restaurant,” Brian Subirana, a research scientist at MIT’s Auto-ID Laboratory, said in a statement.
The research team now plans to incorporate the model into a user-friendly app, which is a non-invasive, free, and convenient tool to help governments detect those potentially infected with SARS-CoV-2, even those not exhibiting any symptoms.
Those who will use the app can get information about whether they might be infected with the virus using an approved diagnostic test. This way, people can have an idea of their risk of being infected.
“AI techniques can produce a free, non-invasive, real-time, any-time, instantly distributable, large-scale COVID-19 asymptomatic screening tool to augment current approaches in containing the spread of COVID-19,” the researchers wrote in the paper.
“Practical use cases could be for daily screening of students, workers, and public as schools, jobs, and transport reopen, or for pool testing to quickly alert of outbreaks in groups,” they added.
Even before the pandemic, scientists have already been studying and training algorithms on cellphone recordings of coughs to accurately diagnose diseases such as asthma and pneumonia. The team believes that the same method could be used to detect COVID-19 in patients.
Eleven months on since the virus was first detected in Wuhan, China, the SARS-Cov-19 pandemic is far from over. Over 46.9 million cases across the globe and 1.20 million people have died. Of these, more than 231,500 deaths have occurred in the United States.
The U.S. reports 9.29 million cases, followed by India, with at least 8.26 million cases. While many countries have skyrocketing cases, some of those who have flattened the curve have resurging infections. The United Kingdom is experiencing a second wave of COVID-19 cases, with a month-long lockdown now planned in an attempt to contain the virus's spread.