DAYTON, Ohio (WDTN) – A University of Dayton Research Institute scientist developed a software that can detect COVID-19 in a matter of seconds.
Barath Narayanan started developing the software to in 2013 to detect lung cancer for his PHD program.
From there, he expanded the software to detect pneumonia, malaria and other cancers.
Narayanan switched gears to use his software to detect COVID-19 when X-rays of COVID-19 patients became available to researchers.
“What we realized is we have data, and we have the experience of working on lung cancer and pneumonia detection, so what if we apply that knowledge and see what we can do from there,” Narayanan said.
The software works by combining X-rays and artificial intelligence.
Narayanan trained the software to learn from markings a radiologist uses to denote positive or negative COVID-19 cases so it could make a diagnosis with accuracy.
A patient would go in for an X-ray as normal and a radiologist would then use the software.
It would use artificial intelligence to examine the X-ray and provide results in a few seconds.
“Like 80% chances, or 90% chances or 0% chances,” Narayanan said. “So it provides a probability number to the radiologist and it could provide him a second opinion and help assist in the decision making process.”
Narayanan adapted the software from its original use in just a few hours.
A few days later, Greenville, South Carolina-based software company Blue Eye Soft reached out to license the software.
“I didn’t expect that to happen at all,” Narayanan said. “I was just doing it for my research, and hopefully to help radiologists, I didn’t think it would happen this quick.”
Currently, the software is waiting on FDA approval.
Once the software is approved and trained on more cases, it will be made available to hospitals and labs across the country.