According to a new study published in the journal Artificial Intelligence can help identify asymptomatic KVID-19 patients only by “forced” cough. IEEE Journal of Engineering in Medicine and Biology. But some experts are skeptical about the technology.
For the study, MIT researchers set up a website where people could record many different pressure cuffs from their cellphones. The researchers collected data from more than 5,300 people. According to a MIT press release, about 2,500 recordings were submitted by people with confirmed COVID-19 diagnoses, including those who tested positive for the virus but were dissatisfied, “2,500 more recordings that they randomly collected from the collection for balance. Selects dataset. ”
The researchers reported that the accuracy rate of the AI model was more than 98 percent when it came to recognizing the presence of COVID-19 in people’s coughs. They also found that technology has effectively differentiated 100 per cent of asymptomatic cough from healthy people.
“MIT researchers have now discovered that people with asymmetry may be different from healthy people with a cough,” the press release said. (Yahoo Life reached out to the main study author but received no response.)
In the statement, Brian Subirana, one of the study’s co-authors, director of MIT’s Auto To-ID Laboratory, says: “We think this shows that you can change the sound when you covid, even if you are asymptomatic.”
While both Dr. Jonathan Chen, Assistant Professor of Medicine at Stanford University Medical Center and Assistant Professor of Clinical Medicine in the Department of Pulmonary, Critical Care and Sleep Medicine at Ohio State University Vaccinar Medical Center. , Tell Yahoo Life that the results of the study are “interesting”, both are questionable.
Chen, who specializes in biomedical informatics research, calls the motivation behind the study “strange” and “clever”, but the “self-report” cough raises questions about the rationale for such success rates. “But a [COVID-19] PCR testing is not so accurate, ”he says. Chen adds that the technology “may still be helpful, but it’s probably much less accurate than what he’s looking for.”
Conroy shares that she has many “serious concerns”, saying: “First, this is not a valid way to investigate the causes of other respiratory disorders or coughs; the study does not include any other respiratory viral infections, including pathology and general. A little worrying for the apparent ability to distinguish between.They further state that the technology can detect the cough of an asymptomatic person infected with COVID-19, but it will be important to note that if you have a new cough, you are not considered asymptomatic. “
She also calls the idea of distinguishing diseases by the sound of coughing “questionable.” Chen adds: “Even if I have a cough, is it a cold?”
According to the publication, the AI model Dell is not designed to diagnose therapeutic people, unless their symptoms are due to COVID-19 or other conditions such as the flu or asthma, adding: “The power of the tool lies in its ability to understand asymptomatic cough from a healthy cough. “
The press release states that researchers are now trying to create an application using technology as a “free, convenient, non-innovative prescreening tool”. “Users can log in every day, cough up their phone and immediately get information on whether they are infected and should therefore be confirmed by a formal formal test.”
However, “Kanroy” strongly “warns” against the use of “unprofessional techniques to guide behaviors.” Instead, she has promoted the “continuous growth of real testing for the SARS-Covy 2 virus, in addition to many of the steps we know to reduce the spread of the virus: universal masking, staying home when you can, avoiding large gatherings and social exclusion.” But the distance when you are around people outside your home. ”
For Latest coronavirus news and updates, Follow along https://news.yahoo.com/coronavirus. According to experts, more than 60 people and those who are immunocompromised are at the highest risk. If you have questions, please refer to this CDCNo and Of the WHO Tool guides.
Read more from Yahoo Life:
Want to get lifestyle and wellness news in your inbox? Sign up here For Yahoo Life’s newsletter.
