Both radiologists and AI models struggle to differentiate between authentic and AI-generated ('deepfake') radiographic images, raising major security and clinical concerns.
Key Details
- 1Research published in RSNA's Radiology shows deepfake X-rays are highly convincing, deceiving even expert radiologists.
- 217 radiologists from 12 centers across 6 countries participated in the study.
- 3Images included AI-generated X-rays (from GPT-4o or RoentGen) mixed with real exams.
- 4Even when aware of the presence of fakes, radiologists could not reliably distinguish between authentic and fake images.
- 5The risk includes fraudulent litigation and potential for clinical harm if synthetic images are injected into hospital records.
Why It Matters
This highlights a major vulnerability in digital radiology workflows, with implications for cybersecurity, clinical accuracy, and legal liability. Increased vigilance and improved detection methods are urgently needed to safeguard medical imaging integrity.
Source
Radiology Business
Related News
•AuntMinnie
Study: Computer Vision Models Best LLMs in Chest CT Breast Abnormality Detection
Computer vision models (CVMs) surpass large language models (LLMs) in accurately labeling incidental breast abnormalities on chest CT scans.
•Cardiovascular Business
Radiology Maintains Lead in FDA-Cleared AI Algorithms, Cardiology Follows
Radiology remains the top specialty for FDA-cleared AI, with cardiology as a strong second, particularly in cardiovascular imaging.
•AuntMinnie
Deep Learning Models Rival Radiologists for Pancreatic Cancer Detection on CT
Deep-learning models achieved comparable or superior accuracy to experienced radiologists in detecting pancreatic cancer on CT scans, especially for small tumors.