Previous Article Next Article  
Researchers Develop AI That Can Predict Lung Disease in COVID-19 Patients
 /  Last Modified: March 31, 2020

This Monday, medical researchers in China and the U.S. report that they have made an artificial intelligence tool that can predict lung disease development in COVID-19 patients.

This is one of the first algorithms during the epidemic that will aid medical workers in prioritizing amid the overloaded health care systems, reports Megan Cofee, physician and New York University’s Grossman School of Medicine professor. She has also co-authored research papers on the new and exciting findings in the ‘Computers, Materials & Continua’ segment at phys.org.

The AI tool discovered an array of factors and instances that are a surefire indicator for the novel infection. Once the tool finds the indicators, it telegraphs strong predictability that the infected patients will develop acute respiratory disease syndrome (ARDS), a common follow-up to the COVID-19 disease. The residual disease is characterized by fluid-flooded lungs and other syndromes, capable of causing fatal complications in more than half of the diagnosed patients.

The researchers responsible for the algorithm collected the data from 53 COVID-19 patients. The patients are mostly from two Wenzhou hospitals in China. They have found that changes in three critical features — liver enzyme alanine aminotransferase (ALT) levels, hemoglobin levels, and reported body aches are the main indicators of an on-set predicted complication.

There are COVID-19 related characteristics that contrast with this research, such as lung images that show “ground glass opacity”, very strong immune responses, and high fevers. Unfortunately, they are not enough to indicate initial ARDS in patients, since they require more factors and variables.

Although further studies show that elderly men at around 60 years are at a higher risk of lung complications, age and sex are not considered as key factors for the predictions.

“It’s been fascinating because a lot of the data points that the machine used to help influence its decisions were different than what a clinician would normally look at,” sais Coffee to AFP, emphasizing the unorthodox approach for the research.

Algorithms and AI in medicine are not unusual, as there are AI-based tools for skin cancer prediction which dermatologists commonly use.

“This is significant because doctors are learning about COVID-19 on-the-fly, can help doctors decide — diagnostically — which way to go, and gives them more criteria upon which to decide which patients should take priority when hospitals around the world are overwhelmed from the COVID-19 pandemic,” says Anasse Bari, a computer science professor at NYU.

It’s imperative that we follow strict protocol and the guidelines of isolation and social distancing, as the pandemic stretches on to all corners and branches of the world.

0 article views
Get our news on Facebook
Join our Group
SCIENCE HAWK
DAILY EDITION
FREE TO YOUR INBOX