A South Korean study finds that AI-generated chest x-ray reports are nearly as clinically acceptable as radiologist-written reports under standard criteria.
Key Details
- 1AI-generated and radiologist-written reports showed similar acceptability under a standard criterion: 88.4% vs 89.2% (p = 0.36).
- 2Under a more stringent criterion (acceptable without revision), AI was less acceptable: 66.8% vs 75.7% (p < 0.001).
- 3The model (KARA-CXR) was trained on 8.8 million chest x-rays from 42 institutions across South Korea and the U.S.
- 4AI-generated reports demonstrated higher sensitivity for referable abnormalities (81.2% vs 59.4%) but lower specificity (81% vs 93.6%) compared to radiologists.
- 5Seven thoracic radiologists independently evaluated report acceptability; most felt AI was not yet ready to replace human radiologists.
- 6Editorials note AI is diagnostically positioned between residents and board-certified radiologists.
Why It Matters
This study demonstrates that AI-generated reporting can meet foundational quality standards, highlighting its potential for expediting workflow in busy or resource-constrained environments. However, the nuanced limitations under more stringent criteria suggest further development is needed for AI to match board-certified radiologist standards.
Source
AuntMinnie
Related News
•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
ChatGPT-5 Performs Poorly at Detecting Distal Radius Fractures on X-rays
ChatGPT-5 missed about half of distal radius fractures on x-rays, showing limited diagnostic capability in a real-world scenario.
•AuntMinnie
In-House LLM Accurately Flags Critical Pediatric Radiology Findings
A locally deployed LLM system at Stanford Health achieved 99.7% sensitivity for detecting critical results in pediatric radiology reports, supporting workflow automation.