AI reduced CT brain reporting time in teleradiology

Thursday, November 30 | 10:10 a.m. - 10:20 a.m. | R3-SSNR15-5 | S402

An AI algorithm deployed to assist with noncontrast CT brain reports in Australia cut radiologist reporting time by up to 8.5%, according to this scientific session.

During the talk, presenter Catherine Jones, MD, of Monash University, the University of Sydney, and the I-MED Radiology Network in Australia, will discuss the evaluation and impact of commercial AI tool within a large teleradiology service. Initially, the AI was deployed in the background to confirm successful technical deployment, with AI findings not made available to radiologists. A full launch followed, and all radiologists were provided access to AI findings as part of adopting the use of the tool as standard practice.

To test report time using the AI algorithm, deidentified details of reporting radiologist, study ID, AI findings, AI access per case, series count, and system reporting time stamps were retrospectively collected for all standalone noncontrast CT brain exams reported between February 2022 and January 2023. Reporting time was defined as the time between the “image-opened” and “dictation-end” time stamps.

A total of 18,550 studies reported by 30 radiologists were included in the analysis. Use of the AI and reporting out of hours was associated with an 8.5% and 7.1% reduction in reporting time, respectively, according to the researchers, who noted that analysis was restricted to studies reported by radiologists who reported a minimum of 50 studies with and without AI assistance.

Even accounting for known factors that affect report time, the availability of a comprehensive AI tool to assist detection of findings has a striking association with reduced time to report noncontrast CT brain cases, indicating high likelihood of improved clinical efficiency in the real-world, teleradiology setting, as Jones will further explain at this Thursday session.


Page 1 of 2
Next Page