UC San Diego Health developed and applied an artificial intelligence algorithm to more than 2,000 lung X-ray images, helping radiologists more quickly identify signs of early pneumonia in COVID-19 patients.
The UCSD Health team partnered with Amazon Web Services for the clinical research study, which aims to quickly detect pneumonia and better distinguish between COVID-19 patients likely in need of more supportive care in the hospital and those who could be monitored at home. The technology allows physicians and radiologists to get an initial estimate of the patient's likelihood of having pneumonia within minutes.
In one case example, an emergency department patient who did not have any symptoms of COVID-19 had a chest X-ray taken for other reasons, yet the AI analysis of the imaging showed early signs of pneumonia. UCSD Health clinicians then tested the patient for COVID-19 and the result came back positive.
"We would not have had reason to treat that patient as a suspected COVID-19 case or test for it, if it weren't for the AI,” said Christopher Longhurst, MD, CIO and associate CMO at UCSD Health, according to the news release. "While still investigational, the system is already affecting clinical management of patients."
UCSD Health radiologists developed the new AI capability several months ago. The clinical team trained the machine-learning algorithm with 22,000 notations by human radiologists so it can overlay X-rays with color-coded maps that indicate pneumonia probability. The health system emphasized they are not diagnosing COVID-19 itself by lung imaging because several types of bacteria and viruses can cause pneumonia.
The AWS Diagnostic Development Initiative donated service credits to help support the clinical study, which has been deployed across UCSD Health. AWS was able to help UCSD Health's clinical research IT team to launch the study system-wide in 10 days, according to the report.