10 May AI for detection of COVID-19 related coughs
| Approach or solution |
|---|
| AI model to detect asymptomatic Covid-19 infections through cellphone-recorded coughs. An algorithm developed by a team of researchers at MIT has correctly identified people with Covid-19 only by the sound of their coughs. It is claimed that the crucial difference in the sound of an asymptomatic-Covid-patient cough could not be heard by human ears. The team is working on embedding this model into an app which if approved by the FDA could turn into a non invasive screening tool to detect people that could be asymptomatic . |
| Organisation or initiative |
|---|
| MIT Auto-ID Laboratory |
| URL or reference |
|---|
| https://news.mit.edu/2020/covid-19-cough-cellphone-detection-1029 |
| Summary of the innovation |
|---|
| Researchers at MIT developed an AI model that can distinguish the caughts of a healthy person, an asymptomatic person infected with COVID and a symptomatic one . Researchers developed their model by analysing samples with recorded cough from tens of thousands of people who send them voluntarily through the phone or web browser. The model correctly identified 98.5% of coughs confirmed as covid-19 and 100% for the asymptomatic ones. When the AI model is fed a cough from an asymptomatic person, is able to detect patterns based on 4 biomarkers : strength , sentiment , respiratory performance and muscular degradation that are specific to COVid-19. The cough of a symptomatic person it’s harder to distinguish. |
| Use cases supported |
|---|
| Screening tool for people who may not suspect they are infected |
Sorry, the comment form is closed at this time.