Some of today’s large language models may be more effective communicators than providers when it comes to explaining interventional radiology procedures. 

The U.S. Food and Drug Administration highlighted concerns about "data drift," which can potentially lead to performance degradation, bias or reduced reliability. 

Just because the FDA has approved an AI-equipped medical device doesn’t mean the agency knows how the techno-apparatus will perform in the real world.

The New York-based radiology vendor announced the news Tuesday, with it applying to a novel product that flags a wide array of life-threatening, time-sensitive conditions in CT scans. 

The tool can identify small focal cortical dysplasia on imaging that is typically missed by radiologists, bottom-of-sulcus dysplasia in particular.

Researchers with the U.S. FDA recently assessed the impact of one such product for pulmonary embolism, with mixed results depending on the time of day. 

Healthcare Interactive confirmed the cyberattack in a statement released last week; however the details have yet to be reported to HHS. The incident happened in July. 

Thanks to AI, clinicians can use mammograms to do a lot more than identify signs of breast cancer. Researchers explored data from nearly 50,000 patients, presenting their findings in Heart.