Researchers at the University of Waterloo have developed an AI-based software tool aimed at improving COVID-19 screening.
Alexander Wong, a system design engineering professor at UW, launched the Covid-Net project in conjunction with DarwinAI — a Waterloo startup that he co-founded. The project also involved researchers from Waterloo as well as radiologists at the Stony Brook School of Medicine and the Montefiore Medical Center in New York.
“This software has had promising initial results,” said Wong. “We hope that by making this software open, we can attract clinicians and scientists far and wide to improve upon the technology.”
The initiative was first launched at the start of the pandemic back in March 2020. The research team talked to clinicians to figure out ways they can help tackle COVID-19 from an engineering perspective.
According to Yahoo News, the technology can assess the severity of COVID-19 among patients with a high degree of accuracy.
The open-source software implements deep-learning AI to screen chest x-rays and CT scans for any signs of coronavirus infections. It was developed to analyze the extent and opacity of infections in the lungs of COVID-19 patients, as stated in the Yahoo News article.
The AI-assessment was then compared with the analysis done by expert radiologists on the same X-ray. Predictions of the AI were on par with the ones provided by the experts.
In a UW press release, in order to develop the screening tool, the AI model was trained from a publicly-sourced dataset compiled from almost 6,000 chest radiography images from nearly 3,000 patients.
“Chest radiography can be conducted quickly and is relatively low-cost and widely available. It is already used by several countries to complement PCR tests,” Wong said. “Augmenting it with AI to improve screening accuracy could have a lot of value.”
Wong also mentioned that the software can provide doctors with an important tool to help manage COVID-19.
“The promising results in this study show that artificial intelligence has a strong potential to be an effective tool for supporting frontline healthcare workers in their decisions and improving clinical efficiency, which is especially important given how much stress the ongoing pandemic has placed on healthcare systems around the world,” Wong said.
In the near future, Wong said he would like to see this technology integrated in hospitals. He added that the AI does not replace healthcare workers — instead, it is a tool that will enhance their workflow by providing them with a better understanding of patients’ conditions.
“As researchers, we’re always looking for ways to give back to the community,” Wong said. “We hope making the AI models and datasets available will drive global innovation in healthcare.”