Johns Hopkins University has highlighted research that shows AI can spot COVID-19 in lung ultrasound images.
In a study published March 11 in Communications Medicine, researchers describe the development of an AI algorithm that analyzes lung ultrasound images to spot features known as B-lines, which appear as bright, vertical abnormalities and indicate inflammation in patients with pulmonary complications.
“The findings culminate an effort that started early in the pandemic when clinicians needed tools to rapidly assess legions of patients in overwhelmed emergency rooms,” the university said, in a news release.
Significantly, the training of the AI involved combining computer-generated images with real ultrasounds of patients, including some who sought care at Johns Hopkins, the release noted.
“Early in the pandemic, we didn’t have enough ultrasound images of COVID-19 patients to develop and test our algorithms, and as a result, our deep neural networks never reached peak performance,” said first author Lingyi Zhao, PhD, who developed the software during postdoctoral fellowship work.
To that end, the group trained the AI to learn from a mix of real and simulated data and then discern abnormalities in ultrasound scans that indicate a person has contracted COVID-19. The tool is a deep neural network, a type of AI designed to behave like the interconnected neurons that enable the brain to recognize patterns, understand speech, and achieve other complex tasks, Zhao noted.
The full study is available here.
