AI Transforming Radiological Science

May 26, 2022

As the pandemic has super-charged the rise of AI-powered medical imaging, a leader in the field, Maryellen Giger, A.N. Pritzker Distinguished Service Professor of Radiology at the University of Chicago, undertook C3DTI’s challenge in March 2020 to apply her considerable expertise to COVID-19 mitigation efforts.

Among developing AI methods for COVID-19 diagnosis and prognosis, the team developed a patent-pending cascaded transfer learning technique to identify with 85 percent accuracy CT scans of COVID patients who should receive steroid treatments. They then applied the same technique to cardiac patients to detect brain injuries earlier than current methods – with broader potential still for using AI to inform clinical decision-making and monitoring patient treatment for a range of conditions.

“Access to large, well-annotated data sets remains the number one requirement for advancement of AI in radiological sciences,” wrote Giger in a 2020 review of the field. Along came COVID, then the C3DTI grant, then follow-on NIH funding, and Giger launched the Medical Imaging and Data Resource Center later that year, expected to be “the largest open database of anonymized COVID-19 medical images, along with associated clinical data, in the world,” according to HealthCareBusiness News.

In January 2021, the Center released their first set of images, providing “a common framework to enable technological advancements, guide researchers’ validation and use of AI, and translate clinical systems for the best patient management decisions,” according to

“Our DTI research prepared us well, and led efforts to create the MIDRC,” says Giger.