British and American scientists have teamed up to develop an artificial intelligence (AI) diagnostic tool that can predict whether a person may be infected with new coronary pneumonia based on symptoms, according to the study published in Nature. Journal of Medicine.
Using data from the “New Coronary Pneumonia Symptoms Study” app, the AI model, which compares people’s symptoms with the results provided by traditional new coronary pneumonia tests to predict whether someone is likely to be infected with new coronary pneumonia, is expected to help people who have difficulty getting tests, and are about to conduct two clinical trials in the U.S. and the U.S., the researchers said.
More than 3.3 million people worldwide have downloaded the “New Coronary Pneumonia Symptoms Study” program and use the app daily to report on their health. The researchers analyzed data collected by the app from 2.5 million U.S. and U.S. users who regularly recorded their health status in the app, with about a third of them recording symptoms associated with neo-coronary pneumonia, and 18,374 reporting that they had been tested for coronavirus, 7,178 of whom tested positive.
Using the data, the team looked at which known symptoms associated with neo-coronary pneumonia were most likely to be associated with positive test results. They found that symptoms of new coronary pneumonia were more common than those of colds and flu. They warn that not just focusing on fever and cough, but that loss of taste and smell (smell disorder) is more of a concern, as two-thirds of people who test positive for coronavirus infection report symptoms. This suggests that loss of sense of smell is a more powerful indicator of neo-coronary pneumonia than fever.
The researchers then created a mathematical model that predicted a person’s likeability to contract new coronary pneumonia, based on a combination of the user’s age, gender, and four main symptoms — loss of smell or taste, severe or persistent coughing, fatigue, and loss of appetite. The researchers applied the model to a group of 800,000 app users with symptoms, predicting that less than one in five (17.42%) of those who were unwell at the time might have contracted new coronary pneumonia.
The researchers believe that combining this AI prediction model with the “New Coronary Pneumonia Symptomstudy Study” can identify people who may be infected when symptoms first appear. Tim B. Professor Spector said: “Taste or loss of smell is a key early warning sign of infection with new coronary pneumonia and should be included in the routine screening of the disease, where anyone who suddenly loses his sense of smell or taste should assume that they have been infected and observe local self-isolation measures.” “