Research Roundup: SFU researchers develop AI tool to better diagnose COVID-19 cases

The researchers are working with Providence Health Care to create the diagnosis tool

Photo courtesy of SFU News

Written by: Nathaniel Tok, Peak Associate

SFU Researchers and Providence Health Care (PHC) are using Artificial Intelligence (AI) to better understand how to detect COVID-19 cases.

The tool, which is currently under testing in St. Paul’s Hospital, will help clinicians distinguish between COVID-19 pneumonia and non-COVID-19 pneumonia. 

It works together with other diagnosis tools like Computed Tomography (CT) scans by analyzing a chest X-ray image to help confirm if the pneumonia’s characteristics are consistent with COVID-19.

Yağız Aksoy, an Assistant Professor in the School of Computing Science, helped create the machine’s learning algorithms that enables the AI tool to analyze X-ray images. “Instead of doctors checking each X-ray image individually, this system is trained to use algorithms and data to identify it for them,” explained Askoy. 

According to Askoy, this will allow resident and less experienced doctors to be able to identify COVID-19 cases even if a senior doctor is not present. 

SFU mathematician Vijay Naidu also helped to create a COVID-19 identifiers database used by the AI tool to better identify positive patients. 

The tool’s development highlights the importance of collaboration during the COVID-19 pandemic between researchers in academia and clinicians in healthcare according to Soyean Kim, director, Digital Products, Providence Health Care.

The tool is still under testing and evaluation, however once it is approved, it will be made available for free with the UN’s support. Fred Popowich the scientific director of SFU’s Big Data Initiative explains “Our goal is to advance COVID-19 response efforts and make this knowledge accessible to clinicians around the world.”