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
MRI AI for Parkinsonian Syndrome Gains FDA De Novo Clearance
Neuropacs Corp.'s MRI AI software earns FDA De Novo classification to assist in Parkinsonian syndrome diagnosis.
•AuntMinnie
Highlights from Recent AI Research in Digital X-Ray Imaging
AuntMinnie Digital X-Ray Insider covers the latest AI advancements and challenges in x-ray imaging.
•AuntMinnie
AI Model Accurately Estimates Bone Density on Pediatric Chest X-rays
A deep-learning AI model accurately estimates bone mineral density using pediatric chest x-rays, showing potential for opportunistic bone health screening.