Part of the challenge of controlling the coronavirus pandemic is to quickly identify and isolate infected people, something that is not easy, considering that the symptoms of COVID-19 are not always obvious, especially in the beginning. Now scientists have developed a new artificial intelligence model that can detect the virus from a simple forced cough. Evidence shows that AI is able to detect differences in cough that cannot be heard with the human ear, something that could become a useful tool for early detection of the disease.
Objective: stop the spread of COVID-19
Scientists have been investigating alternative ways of detecting the disease, such as thermal cameras to help detect anyone with a fever, for example, but we know that this is not the only telltale sign of infection that we can observe.
A team of researchers from the Massachusetts Institute of Technology (MIT) has developed an AI that can detect people infected with COVID-19 by another of its main symptoms: cough. The virus causes a persistent cough and there is evidence that it can change the way we make the sound even if we do not have symptoms.
In the experiment, the algorithm achieved a 98.5% success rate among people who had received an official positive coronavirus test result, increasing to 100% in those who had no other symptoms.
“The way sound is produced changes when you have Covid, even if you don’t have symptoms,” explains MIT scientist Brian Subirana, co-author of the study published in the IEEE Journal of Engineering in Medicine and Biology.
It is an incredible result, of course, but there is still a lot of work to be done. The researchers emphasize that its main value lies in detecting the difference between healthy cough and unhealthy cough in asymptomatic people, not in the actual diagnosis of COVID-19. It is an early warning system.
When could it be used?
Regulatory approval is required to develop it into an application. Researchers continue to test their AI in clinical trials and have already begun applying for Food and Drug Administration (FDA) approval for use as a screening tool.
The algorithm builds on previous models the team developed to detect conditions such as pneumonia, asthma, and even Alzheimer’s disease. Among its strengths is that the test is non-invasive, practically free, and can be applied quickly.
We have already commented that many companies have already installed heat cameras in their entrances to detect people with a fever, so this new technology could be used in a similar way. It can detect audio signals from an infected person much more accurately than a human could, and in some cases it could detect asymptomatic people.
The study was supported by the pharmaceutical company Takeda Pharmaceutical Company Limited.
Referencia: COVID-19 Artificial Intelligence Diagnosis using only Cough Recordings. IEEE Open Journal of Engineering in Medicine and Biology 2020 DOI: 10.1109/OJEMB.2020.3026928
