SFU and Providence Health Care collaborate on artificial intelligence project to quickly diagnose COVID-19

    1 of 2 2 of 2

      The advancement of artificial intelligence and machine learning has alarmed some high-profile billionaires.

      Among them are Mark Cuban, who worries about the creation of a real-life Terminator, and Elon Musk, who once likened it to "summoning the demon".

      Even Microsoft founder Bill Gates has expressed serious concerns in the past.

      But there are also many upsides to machine learning.

      For instance, Simon Fraser University researchers and Providence Health Care officials are making use of it in the fight against COVID-19.

      They're collaborating on creating a new AI tool to diagnose the presence of the novel coronavirus.

      It's in the process of being validated at St. Paul's Hospital in Vancouver's West End.

      “Instead of doctors checking each X-ray image individually, this system is trained to use algorithms and data to identify it for them,” SFU assistant computing science professor Yağiz Aksoy said in a university news release.

      MAGPIE Group postdoc Vijay Naidu and SFU computing science assistant professor Yağız Aksoy played key roles in developing the system.

      This is accomplished by enabling clinicians to run X-ray images through a computer to determine if a positive test for pneumonia is consistent with that which occurs in COVID-19 patients.

      The system was developed by Aksoy and MAGPIE Research Group postdoc mathematician Vijay Naidu. They rely on machine learning to identify unique characteristics found in the virus.

      This can allow for speedier diagnoses by health-care professionals on the frontlines before the patient is assessed by a senior physician.

      “Connecting partners to diverse SFU experts is core to SFU’s Big Data Initiative,” Fred Popowich, scientific director, SFU's Big Data Initiative, said in the news release. “Our goal is to advance COVID-19 response efforts and make this knowledge accessible to clinicians around the world.”

      To keep people from recklessly self-diagnosing, the Straight is partnering with , a destination for digital health education. Its platform has information curated and collected from more than 60 reputable health association partnerships and the Mayo Clinc.