A novel hybrid deep learning framework based on biplanar X-ray radiography images for bone density prediction and classification
Authors' conclusion
This study demonstrates that deep learning neural networks applied to BPX images can accurately predict BMD and perform classification diagnoses, which can reduce the radiation risk, economic consumption, and time consumption associated with specialized BMD measurement.
Study context
- Population
- Subjects with BPX scans from Huashan Hospital Medical Checkup Center
- Participants
- 453 subjects906 BPX scans
- Comparison
- Hybrid deep learning framework (HDLF) models based on BPX images vs BPX images plus clinical information (multimodal), against quantitative computed tomography (QCT)
- Outcome
- Bone mineral density (BMD) prediction in the independent test set
Methodological detailsdeep learning (hybrid deep learning framework), training-validation set:independent test set ratio 4:1, QCT as reference standard
- Deep learning (hybrid deep learning framework)
- Training-validation set:independent test set ratio 4:1
- QCT as reference standard
Performance
As reported in the abstract
| Test or model | R 2 | Pearson correlation coefficient |
|---|---|---|
| BPX images | 0.77 | 0.88 |
| Multimodal | 0.79 | 0.89 |
R 2
BPX images0.77
Multimodal0.79
Pearson correlation coefficient
BPX images0.88
Multimodal0.89
Further findings
| Outcome | Multimodal |
|---|---|
| Classification (abnormal BMD) in the independent test setHighest AUC achieved with multimodal data | |
| AUC | 0.97 |
| Accuracy | 0.93 |
| Sensitivity | 0.84 |
| Specificity | 0.96 |
| F1 score | 0.93 |
Original publication
A novel hybrid deep learning framework based on biplanar X-ray radiography images for bone density prediction and classification. Osteoporos Int. Published 15 January 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →