Radiology AI devices are more likely to face FDA recalls, largely due to deviations from intended use and incomplete clinical data.
Key Details
- 1New analysis looked at 903 healthcare AI devices approved by FDA between 1995 and 2024.
- 2Of these, 5% (43 devices) were recalled, with a median recall interval of 458 days.
- 3Radiology comprised 692 of these devices, with a 4.3% (30 devices) recall rate.
- 4Radiology AI devices are 52% more likely to face recall than AI devices in other specialties.
- 5Missing clinical study information and deviations from intended use were major risk factors for recall.
- 6Problems with temperature and labeling led to lower recall risk, while market surveillance flags increased it.
Why It Matters
This study is the first to systematically quantify recall risk factors for AI-enabled devices, highlighting the urgent need for rigorous validation and postmarket surveillance, especially within radiology. Increased transparency and robust premarket evaluation may help mitigate safety risks as AI adoption accelerates.
Source
Radiology Business
Related News
•AuntMinnie
AI’s Path in Radiology Hinges on Tech, Policy, and Human Insight
SIIM-CAIMI keynote highlights that AI's future in radiology depends on synchronized advances in technology, regulation, finance, and culture.
•Radiology Business
Meta Funds Ohio State’s Radiology AI Innovation Projects
Meta will fund Ohio State researchers to explore AI-driven advancements in radiology, including chest X-ray reporting and disease prediction.
•AuntMinnie
SIIM-CAIMI Leaders Discuss Key Trends in Imaging AI Deployment
SIIM-CAIMI program chairs highlight AI's transition from algorithm evaluation to real-world implementation in medical imaging.