A deep learning model accurately identifies hard-to-detect physeal fractures in children's wrist x-rays.
Key Details
- 1AI model developed using 2,103 x-rays from 1,082 pediatric patients (mean age 10)
- 2Best performing model (EfficientNet-B2) achieved 84% accuracy, 81% precision, 89% recall, and F1-score of 0.86 on test data
- 3Model uses Grad-CAM for interpretability, highlighting regions contributing to decisions
- 4Physeal fractures are often misdiagnosed (up to 46% rate), risking later growth deformities
- 5Proposed use as decision-support tool in urgent care and ER settings; potential to reduce missed diagnoses
Why It Matters
Missed pediatric fracture diagnoses can cause significant long-term harm. AI decision-support could help frontline clinicians recognize subtle injuries and improve outcomes, addressing a known gap in pediatric musculoskeletal radiology.
Source
AuntMinnie
Related News
•Radiology Business
NYC Health + Hospitals CEO Considers AI to Replace Radiologists
NYC Health + Hospitals CEO suggests AI could partially replace radiologists, pending regulatory approval.
•AuntMinnie
AI Models Reveal Racial Disparities in Breast Cancer Patterns
Machine learning models reveal significant racial disparities and key predictors in breast cancer incidence across diverse groups.
•Radiology Business
UCLA Appoints Inaugural Associate Dean for Health AI Strategy
UCLA has appointed Katherine P. Andriole as its first associate dean for Health AI Strategy and Innovation, with an initial focus on radiology.