Purpose
Missed fractures are the primary cause of interpretation errors in emergency radiology, and artificial intelligence has recently shown great promise in radiograph interpretation. This study compared the diagnostic performance of two AI algorithms, BoneView and RBfracture™ version 2.2, in detecting traumatic abnormalities (fractures and dislocations) in MSK radiographs.
Key findings
- No statistically significant differences were found in sensitivity or specificity between BoneView and RBfracture
- The ratio of subtle to obvious findings in missed abnormalities was not statistically significantly different between AI algorithms
- McNemar’s tests comparing BoneView and RBfracture indicated no significant differences in sensitivity or specificity in any of the included MSK regions
- Performance was similar in adults and children
- RBfracture demonstrated a sensitivity of 87.2% and a specificity of 89.2%
- RBfracture demonstrated PPV of 85.1% and NPV of 90.8%
- RBfracture demonstrated an accuracy of 88.4%
Open access
This research is open access
This original research can be read in Emergency Radiology, published in June 2025.
Funding and conflicts of interests
The authors have no conflicts of interest to declare that are relevant to the content of this article.
