Research finds insufficient data transparency for FDA-cleared AI devices in Alzheimer’s imaging, raising bias and generalizability concerns.
Key Details
- 124 AI/ML-based devices for Alzheimer's disease and related dementias (ADRD) were FDA-authorized between January 2015 and December 2024.
- 291.7% (22/24) were cleared via the 510(k) pathway; 8.3% (2/24) through de novo classification.
- 3For 12 devices, there was no info on training or validation data in FDA summaries or peer-reviewed articles.
- 4Only 2 FDA summaries (8.3%) included external validation data; 10 peer-reviewed articles (41.7%) provided such data.
- 5Reporting of disease status, age, sex, race, and ethnicity was rare across most devices; demographic transparency was lacking.
- 6Most devices (21) were reviewed by radiology panels and primarily involved MRI and PET imaging.
Why It Matters
Transparent disclosure of training and validation data, especially demographic details, is vital for assessing bias and clinical accuracy in AI-based imaging devices. Without sufficient data transparency, there is uncertainty about the generalizability and safety of these AI tools in real-world, diverse patient populations.
Source
AuntMinnie
Related News
•AuntMinnie
Paul Chang Discusses Foundation Models and Agentic AI at RSNA 2025
Dr. Paul Chang shares his insights on the role of foundation models and agentic AI in radiology at RSNA 2025.
•AuntMinnie
Toronto Study: LLMs Must Cite Sources for Radiology Decision Support
University of Toronto researchers found that large language models (LLMs) such as DeepSeek V3 and GPT-4o offer promising support for radiology decision-making in pancreatic cancer when their recommendations cite guideline sources.
•AuntMinnie
AI Model Using Mammograms Enhances Five-Year Breast Cancer Risk Assessment
A new image-only AI model more accurately predicts five-year breast cancer risk than breast density alone, according to multinational research presented at RSNA 2025.