LONDON (Commonwealth Union) – The use of Artificial Intelligence (AI) in predicting patterns and behaviors has become a common practice in recent years.
A new study conducted by University College London researchers have found that a blood test taken during a COVID-19 infection could forecast who has a greater chance of developing long COVID. The study evaluated proteins in the blood of healthcare workers who had SARS-CoV-2, comparing them to samples from healthcare workers who were uninfected.
Protein levels in the body are generally stable, but the researchers saw a significant change in levels of some of the proteins up to 6 weeks following infection, indicating disruption to a number of significant biological processes. AI algorithm, assisted in identifying a “signature” in the abundance of varieties of proteins that successfully forecasted the possibility of a person reporting persistent symptoms a year after infection.
The researchers indicated that, if these results are repeated in a bigger, independent group of patients, a test could possibly be offered alongside a polymerase chain reaction test that could forecast an individual’s likelihood of having long COVID.
“Our study shows that even mild or asymptomatic COVID-19 disrupts the profile of proteins in our blood plasma. This means that even mild COVID-19 affects normal biological processes in a dramatic way, up to at least six weeks after infection,” said Dr Gaby Captur who was the lead author. She also indicated that tool of predicting long COVID needs validation in an independent, larger study with more patients but stated that using their technique will give a test that forecasts long Covid at a time of initial infection and could see a speedy implementation which would also be economical.
The researchers saw that at the time of initial infection, abnormal rates of 20 proteins evaluated could forecast continuing symptoms after a year. Most of these proteins were connected to anti-coagulant and anti-inflammatory activity.
