Tuesday, December 3 | 8:30 a.m.-8:40 a.m. | T1-SSCH05-4 | E451A
For this session, researchers will highlight a significant reduction in radiologists' reading time from the use of AI software for lung nodule detection on CT exams.
The team from Jefferson Health and Thomas Jefferson University (TJU) Hospitals in Philadelphia led an institutional review board (IRB)-approved pilot research study using InferRead Lung CT AI software (Infervision). They evaluated the interpretation times of nine community-based radiologists before and after deployment of the PACS-integrated AI software over a 90-day period.
During the trial period, 1,192 chest CTs were interpreted without the detection tool, and 738 scans were read with it. Reading time distributions will be shared during the session.
In advance, however, presenter Baskaran Sundaram, MD, professor in thoracic radiology and director of the division of cardiothoracic radiology at TJU, and colleagues noted reading time was 16.5 plus or minus 11.4 minutes without AI, with the peak time between 8-10 minutes, while the reading time was 14.8 plus or minus 10 minutes with AI and a peak time between 6-8 minutes.
The deployment of the tool not only significantly reduced reading times but also bolstered radiologists’ confidence in reporting lung nodules, diminished anxiety, and enhanced the overall reading experience, according to the researchers.
Were results the same across all nine hospitals? Attend to find out.