Vancouver-based radiologists are at the helm of an international study that aims to better predict the presence of COVID-19 using CT scans.
“We know the lungs of COVID-19 patients are white and hazy, like a white-out or blizzard,” said Dr. Savvas Nicolaou, who is heading the research.
“Currently, we can’t predict disease severity and its clinical impact in different patient populations. We’re confident this new tool will help us do that,” said Nicolaou, director of emergency and trauma radiology at VGH.
Health-care workers have seen patients in emergency who have non-typical symptoms, such as severe abdominal pain and stroke and acute chest pain. Upon reviewing those patients’ CT scans for those conditions, the “tell-tale haziness” of COVID-19 was seen in their lungs, Nicolaou explained.
Clinical data from the study will be integrated into a model that could help physicians determine whether individuals are best treated at home or whether they may require hospitalization and/or ventilation.
It could also help flag individuals who may ultimately develop permanent lung damage or fibrosis.
“The model will also assist in detecting similarities and differences in variations of patterns across different cultural and ethnic groups,” said Dr. Kendall Ho, VGH emergency physician and academic director of the UBC Cloud Innovation Centre. “(It will also) help us understand early and late stages of patters of disease.”
Once developed, the new AI model will be piloted at VGH, with an aim of utilizing it in routine diagnostic procedures to improve the accuracy of COVID-19 diagnostics.