Nat Med: Developing a new AI diagnostic tool to predict the risk of COVID-19 infection without testing

2020 May 17 News /BioValleyBIOON / - Recently, research onNatureMedicinean international journal published report, scientists from the University of London, King's College and other institutions through research and development of artificial intelligencea diagnostictechnology that can help predict the likelihood of COVID-19 according to the body of an individual suffering from the symptomsThis model can AI with data from studiesappCOVID symptoms of, an individual by comparing the results of detection of the symptoms and the conventional COVID to predict a person's risk COVID-19, which can help detect indeed limited population screening COVID-19, now about to begin clinical trials in twothe United Kingdom and the United StatesSource:CC0PublicDomainmore than 330 million people worldwide have downloaded the app, and use it every day to record their own health status, such as whether they feel good or what new disease symptoms, such as persistent cough, fever, fatigue and loss of taste and so onArticle, the researchers recorded their own health status regularly inapp , the data from the United Kingdom and the United States 2.5 million people were analyzed, of which about one-third of the population recorded with COVID-19 related symptoms, there are 18,374 individuals indicated that they had received to detect coronavirus, there are 7178 people a positive result The researchers also investigated what COVID-19 associated with the body's symptoms may be associated with a positive result, they found that compared with colds and flu symptoms COVID-19 infection of more extensive and researchers to remind you not only concerned with fever and cough; in fact, they found that patients with loss of taste and smell is also very important symptoms of infection, of which two-thirds of coronavirus infection on the report of the symptoms, and in the test results were negative the participants, just over one-fifth of the population reported symptoms; the results show that, compared with fever, loss of taste and smell are predictors of a stronger COVID-19, which supports the sense of smell and taste loss is a common symptom of the disease The researchers then developed a mathematical model to accurately predict the performance of 80%, which can be based on the individual's age, sex and a combination of four key symptoms (including loss of sense of smell or taste, or severe persistent cough, fatigue and not eating) to accurately predict whether an individual has COVID-19, this model will be applied to the entire 800,000 app user groups, which will be able to predict with fewer than one in five (17.42% ) of the population may suffer from COVID-19 The researchers pointed out that the wide spread of AI prediction model and app combination or to help identify the symptoms begin to appear once the crowd is likely to be infected with SARS-CoV-2, but also can effectively track and to monitor their infection last researcher Professor TimSpector said, results of this study showed that the loss of smell or taste is the key to early signs COVID-19 infected with the disease, it seems that the latter should be incorporated into routine screening program, the researchers who strongly urged governments and local health departments should widely publicize this information, and suggested that any people have experienced a sudden loss of sense of smell or taste to detect and take action consciously isolated (Biovalley Bioon.com) Original source: CristinaMennietal, Real-timetrackingofself-reportedsymptomstopredictpotentialCOVID-19 , NatureMedicine < br /> (2020) .DOI: 10.1038 / s41591-020-0916-2