ALBUQUERQUE, N.M. (KRQE) – As a result of the pandemic, hundreds of medical papers related to COVID-19 are being published every week. A University of New Mexico scientist believes there’s a better way to sort through all of the data by using technology to do what humans can’t do.
The COVID-19 pandemic has highlighted the need to share data, sort through misinformation, and find answers; scientific research is expedited.
“We currently have hundreds of publications related to COVID that are coming out on a daily basis,” said Dr. Tudor Oprea, Scientist and Professor of Medicine at UNM Health Sciences Center. “No single human mind can comprehend or read all those publications.”
Oprea believes there’s a way to help the peer review process by using artificial intelligence (AI) and machine learning (ML) to do what humans can’t. An automated system, he says, would be able to comb through millions of pages of text looking for discrepancies, even predict clinical trial outcomes, and alert scientists to unique findings.
“The machine system would actually flag, ‘these observations are consistent with what we’ve seen before, or these observations are surprising, they’re novel, we haven’t seen them before so pay more attention to these,” explained Oprea. “Or better yet, ‘here’s two publications that contradict each other,’ so try to investigate more.”
He said the recent analysis shows roughly 367 COVID-19 papers are being published every week, with a median time from submission to acceptance of just 6 days, compared to the 84-day turnaround time for non-COVID content.
“That is highly irregular and unusual,” said Dr. Oprea. “To my mind, what that really means is that not enough time is given to verify the claims that are made in those publications,” he added.
He’s hopeful AI can help study viral mutations, and identify effective treatments for COVID-19. He envisions scientists being able to access the tool remotely from a cloud server, to operate on a laptop or mobile device.
“We’ve seen enough signal to be confident that a cure exists, it’s just we don’t have it yet. But it’s within reach, we’re getting there,” said Oprea. He said people shouldn’t be afraid of the technology, but rather understand its usefulness.
“Without realizing, we trust computers every day,” said Oprea. “We use Google Maps to get from A to B. This would be the equivalent of doing the same thing, like Google Maps, except going from A to B in scientific space.”
The UNM scientist says he and his colleagues have already been using the same technology in the fights against Alzheimer’s and cancer. He hopes to eventually provide free, open access to the technology to help scientists fight and study COVID-19.
Much of the information published about COVID-19 are opinions, or scientific publications under ‘preprint,’ or unfinished work. “People just want to be heard,” said Oprea.
“My problem with that is that everybody talks, but nobody listens,” said Oprea. “And so we need the computer to start listening to make sense of what’s going on.”
