Tuesday, December 1 | 3:00 p.m.-4:00 p.m. | T7-SSIN04-5 | Room E450B
An AI tool that summarizes patients’ clinical history may save radiologists about 40 minutes a day, according to this scientific presentation.
Synthesizing complex clinical histories from fragmented electronic health record (EHR) data is time-consuming and adds to radiologists’ cognitive burden, noted presenter Abraham Baez Suarez, PhD, of the Mayo Clinic, and colleagues. To help, the group developed GRACIE, a large language model (LLM)-based multi-agent system that presents a structured, time-ordered patient summary at the time of interpretation.
In March and April 2026, the researchers surveyed staff radiologists, fellows, and residents at their institution about adoption, usage frequency, time savings, and workflow impact.
Nearly 93% of respondents actively used GRACIE, and 70% of users relied on it often or always. Users reported median time savings of five minutes per case. Nearly all said it saves time and reduces chart-search burden, and about three-quarters said it uncovers facts they might otherwise have missed. Each summary drew on a median of 109 EHR entries per patient, including lab results, clinical notes, and prior radiology reports.
“Automated, citation-enabled EHR summarization at image interpretation (or procedure) may reduce radiologist cognitive burden, allowing greater focus on primary image analysis,” the researchers wrote.
Grab a seat at this Tuesday session to learn more.



