University of Minnesota and M Health Fairview researchers created a new artificial intelligence algorithm that evaluates chest X-rays to diagnose potential cases of COVID-19 and made it available at scale for free in Epic's App Orchard, according to an Oct. 1 news release.
Four details:
1. The algorithm will be available at no cost to other health systems that use Epic, and current clients can install the tool in as little as 10 days.
2. To develop the algorithm, University of Minnesota and M Health Fairview, both in Minneapolis, analyzed de-identified chest X-rays taken at M Health Fairview since January. The researchers used 100,000 X-rays of patients without COVID-19 and 18,000 X-rays of patients with the virus.
3. The researchers used the X-rays to train the algorithm to diagnose COVID-19; when a patient arrives at the emergency department with symptoms and receives a chest X-ray, the algorithm automatically evaluates the scan and within seconds can notify the care team in the Epic EHR if it recognized patterns associated with COVID-19 in the chest X-ray.
4. After validating the algorithm, the researchers worked with Epic to develop the infrastructure around the algorithm so it could seamlessly translate its findings into the EHR software and notify care teams.