Artificial intelligence is able to identify coronavirus symptoms

Published by:
Posted on: July 21, 2020 7:16 pm EDT

AI and machine learning are making great strides in the field of medicine

So far this year, the world has been dealing with the effects of the coronavirus pandemic that most likely will stay for a while. The biggest challenge when fighting against this virus is the fact that there are many asymptomatic people, as well as those who only demonstrate minor symptoms, so it’s hard to determine the progress of the virus. But, according to a recent British study, artificial intelligence (AI) and machine learning (ML) are being used to overcome this challenge, and a new tool has allowed researchers to group the most common symptoms of the disease.

Through the use of a symptom-tracing app along with ML software, this tool can not only group the most common symptoms, but it can also assess how likely it is for each of those symptoms to develop into a severe case. The app, Covid Symptom Tracker, is proving its worth to serve as a tracker to study how coronavirus cases change among different communities, as well as learning about the symptoms. It was estimated that more than four million people are using this app.

The study used an algorithm for machine learning to explore how likely it was for the symptoms to be clustered together. The study team used data from 1,653 users who tested positive and who had added information about their symptoms and health on a regular basis. Out of the total number of subjects, 383 of those patients visited the hospital at least once and 107 required oxygen or ventilation. As a result, researchers believe that symptoms can be clustered together to predict how likely it is for a severe case to manifest.

Investigators concluded from this study that a higher body mass index, older age and chronic lung disease correlated with more severe symptoms. Also, men seem to report more severe symptoms.