A new AI system is able to identify pneumothorax more accurately than many radiologists.
The system, designed by a team at the University of Waterloo in Ontario, Canada, successfully identifies pneumothorax cases 75% of the time. Specialists, the researchers noted, average diagnosing 50% of pneumothorax cases.
“Our results are very exciting,” Antonio Sze-To, a postdoctoral fellow at the University of Waterloo, said in a news release from the school. “The AI we use works almost like magic—and it will help radiologists save lives.”
The AI system searches through a database of more than 550,000 X-rays when making its analysis. It could make an impact on patient care by helping prioritize certain cases or offering radiologists a second opinion, the researchers noted.
The team behind this new software is working with the Toronto-based University Health Network (UHN) and the Vector Institute to increase its accuracy to more than 90%. Once that milestone is reached, they plan on integrating into the UHN’s Coral Review software system and tracking its success.
“We’re very excited about the collaboration with the University of Waterloo and the Vector Institute,” Leon Goonaratne, senior director at UHN Digital, said in the news release. “It’s an opportunity to further improve the quality process we’ve implemented at hospitals across the province.”
The team’s research is set to be presented at the Vector Institute’s Evolution of Deep Learning Symposium in Toronto.
Michael has more than 16 years of experience as a professional writer and editor. He has written at length about cardiology, radiology, artificial intelligence and other key healthcare topics.