Thursday, December 4 | 8:40 a.m.-8:50 a.m. | R6-SSGU07-2 | Room E353B
In this session, researchers will present results from an international study assessing the performance of prostate MRI AI software in a variety of practice settings.
In their study, presenter Jasper Twilt, a doctoral candidate from Radboud University Medical Center in the Netherlands, and colleagues tested a hypothesis that their AI software could be diagnostically interchangeable with radiology readings for clinically significant prostate cancer across an external, global, multiethnic population.
After training, the software was utilized on over 1,900 patients from population-based screening and primary diagnostic settings and at more than 19 cities in 12 countries in Europe, North and South Americas, Asia, and Australia. The primary end point is the proportion of AI-based assessments that are in agreement with the standard of care diagnoses.
All data will be presented in Chicago.
“AI systems can alleviate the increasing workload, prevent overdiagnosis, and reduce the dependence on expert radiologists across the diagnostic pathway of prostate cancer,” the authors wrote. “If deemed robust across population shifts, geographical sites, acquisition protocols, and the disease spectrum, such systems have the potential to improve workflow efficiency, patient outcomes, and healthcare equity at scale.”
Get all of the details by attending this presentation on the final day of RSNA 2025.
