This AI can identify the cough of asymptomatic people with Covid-19



[ad_1]

Sometimes artificial intelligence knows more about us than we know about ourselves. And when it comes to diagnosing Covid-19, that could be a boon in fighting a pandemic.

Researchers at MIT have developed an algorithm that they say can differentiate the forced coughs of asymptomatic people who have Covid from those of healthy people. They are currently working on a free app that would allow anyone to cough on their smartphone and essentially get a pre-assessment of whether they might have Covid, even if they have no symptoms.

The researchers published their article in the IEEE Journal of Engineering in Medicine and Biology, and MIT News spoke with the team about the investigation to learn more.

Before the pandemic, the MIT team had been working on artificial intelligence models that could diagnose diseases, particularly Alzheimer’s, based on voice recordings. By training the AI ​​on 200,000 recordings of speech and cough, they found that the AI ​​could detect biomarkers such as vocal cord strength, lung capacity, and neuromuscular degeneration that appear in these recordings and differ from the coughs and words of healthy people. .

Since Covid can have some of the same symptoms (in particular, “temporary neuromuscular impairment”), the researchers decided to apply their Alzheimer’s framework to detect Covid-19. The idea is that AI can capture these differences in coughs caused by neuromuscular impairment and other symptoms between healthy people and asymptomatic people with Covid-19 – differences that those asymptomatic people might not notice on their own.

According to his research, the theory worked. Since they had already trained the initial model on tens of thousands of recordings, they were able to create a specific Covid assessment from 4,000 cough samples, half from healthy people, and half from asymptomatic Covid-19 patients.

They then tested the AI ​​on an additional 1,000 samples (again, half healthy, half not), and correctly identified 98.5 percent of coughs from people with Covid-19. It accurately marked all forced coughs from asymptomatic Covid-19 patients. Very impressive!

The researchers say further refinement of the AI ​​model is needed with more training samples. However, they are hopeful that this could be a useful tool to combat the pandemic.

Research shows that asymptomatic Covid patients unknowingly contribute to the spread of the disease. A cough pre-detection tool that allows people to get an easy idea of ​​whether or not they have Covid-19 before they unknowingly go out into the world and potentially infect others could combat that.

Imagine, before you leave the house, asking Alexa to check if you should probably go get a Covid test just by letting Alexa listen for your cough. That scenario may sound a bit strange, but if it works, we will accept it.



[ad_2]