Asymptomatic COVID-19 patients can spread the novel coronavirus without knowing that they are sick.
A recently developed AI with a highly trained “ear” can detect asymptomatic cases of infection with COVID-19 from listening to patients cough and detecting subtle changes that are otherwise inaudible to the regular human year. The researchers are currently field-testing the app in clinical trials. They have already begun requesting approval from the FDA (Food and Drug Administration) for it to work as a screening tool.
How It Works
The algorithm was developed after analyzing past models the team made to discover conditions like pneumonia, Alzheimer’s disease, and asthma.
Apart from memory loss, Alzheimer’s disease can also produce other body degradation like weakened vocal cords and respiratory output.
The Alzheimer’s model was among the most useful resources the researchers used to create the algorithm to detect COVID-19.
Co-author Brian Subirana said:
“The sounds of talking and coughing are both influenced by the vocal cords and surrounding organs.”
“Things we easily derive from fluent speech, AI can pick up simply from coughs, including things like the person’s gender, mother tongue, or even emotional state. There’s, in fact, sentiment embedded in how you cough,” he added.
Performance
The algorithm seems to be working wonders. Testing revealed that it correctly identified 98.5% of patients who have COVID-19 and accurately ruled out infection in 94.2% of the uninfected individuals.
In asymptomatic cases, the app identified 100% of infected people and accurately ruled out COVID-19 in 83.2% of the healthy individuals.
Still, further research is needed to testify that the AI can accurately evaluate coughs from people of all ages and locations.