Transferability of bone phenotyping and fracture risk assessment by μFRAC from first-generation high-resolution peripheral quantitative computed tomography to second-generation scan data
Authors' conclusion
We demonstrated that bone microarchitecture models trained using first-generation scan data generalize well to second-generation scans, performing with a high level of accuracy and reproducibility. Less than 4% of individuals' estimated fracture risk led to a change in treatment threshold, and in general, these dissimilar outcomes using second-generation data tended to be more conservative.
Study context
- Population
- Individuals scanned on both first- and second-generation high-resolution peripheral quantitative computed tomography (HR-pQCT) scanners on the same day
- Participants
- 60 individuals
- Comparison
- First-generation trained bone phenotyping and μFRAC models applied to first- vs second-generation HR-pQCT scan data
- Outcome
- Agreement of first-generation trained models across scanner generations
Methodological detailssame-day scans on both scanner generations, same-day repeat scans for reproducibility
- Same-day scans on both scanner generations
- Same-day repeat scans for reproducibility
Main finding
As reported in the abstract
Bone phenotyping model accuracy across scanner generations
93.1%
5-year fracture risk by μFRAC, Pearson's r across scanner generations (three μFRAC variations)
r > 0.83
Further findings
Reproducibility (same-day repeat scans)
Bone phenotyping accuracy
μFRAC correlation coefficient First-generation 0.99 · Second-generation 0.99
Original publication
Composed by AI from the published abstract and checked against it. How Briefs are made →