Rad AI's software uses machine-learning algorithms to automatically generate "impression" sections of radiology reports that are customized to the preferred language of the radiologist.

Initial user surveys have indicated that the software can lead to significant reductions in error rates and turnaround times and also may reduce radiologist burnout, the company said. Early studies conducted at the Einstein Healthcare Network in Philadelphia and several radiology practices have shown time savings of 20% for interpreting CT scans and 15% for radiographs.

Rad AI will display its AI technology at the upcoming RSNA 2019 meeting in Chicago.