As imaging volumes continue to grow and demands on radiologists increase, new research highlights how AI can support more accurate interpretations of diagnostic quality and reporting efficiency.
In a study of more than 237,000 CT and X-ray exams, Mosaic’s investigational vision-language models (VLMs) demonstrated promising results across 50 conditions evaluated, enhancing detection by approximately 40%.
The research also found AI-generated reports had comparable or lower clinically significant error rates than original radiologist reports. Radiologists using these models saw productivity gains of more than 27% on CT and 31% on X-ray.
Explore all of the findings and what they could mean for the future of radiology.