Artifact-Controlled Multi-Vertebral Transfer Learning Ensemble for Patient-Level Three-Class Osteoporosis Screening Using Thoracoabdominal CT-Derived Lumbar Images
Open Access
- 3 September 2026
- journal article
- research article
- Published by MDPI AG in Biomedicines
- Vol. 14 (9) , 1988
Abstract
Background/Objectives: Osteoporosis and osteopenia remain underdiagnosed despite routine CT examinations containing potentially useful bone information. This study evaluated the feasibility of patient-level three-class classification of Normal, Osteopenia, and Osteoporosis classes using lumbar vertebral PNG images derived from routine thoracoabdominal CT scans, without raw DICOM data, HU values, or quantitative BMD information. Methods: The dataset comprised 190 patients labeled by DXA-derived mean L1–L4 T-scores and 760 vertebral images. Four vertebral images per patient were analyzed jointly using patient-level 5-fold stratified cross-validation. ROI annotations defined only the standardized crop center and were not used as model inputs, masks, or features. ResNet18, ResNet50, and DenseNet121 transfer learning models were trained with patient-level L1–L4 feature fusion, and their probabilities were combined using equal-weight soft voting. Results: Within the patient-level 5-fold cross-validation framework (the same fold was used for model selection and final performance reporting, rather than serving as an independent held-out test set), the final ensemble reached a cross-validation accuracy of 0.7474 (95% CI: 0.6842–0.8053), balanced accuracy of 0.7377 (95% CI: 0.6703–0.8009), macro-F1 of 0.7403 (95% CI: 0.6730–0.8011), and macro-average ROC-AUC of 0.8576 (95% CI: 0.8091–0.9013). ROI occlusion reduced performance, whereas mask-only and geometry-only controls indicated that ROI annotation or geometry alone was insufficient for class discrimination. Grad-CAM showed that activations were not confined to ROI annotations, although residual ROI-related effects could not be fully excluded. Conclusions: Patient-level multi-vertebral transfer learning demonstrated moderate feasibility for image-based osteoporosis decision support, supporting a preclinical decision-support role rather than direct diagnosis or replacement of QCT/BMD assessment.Keywords
This publication has 21 references indexed in Scilit:
- Opportunistic osteoporosis screening using CT-based Hounsfield units: a machine learning approach for preoperative spine surgery planningEuropean Spine Journal, 2026
- Opportunistic screening of osteoporosis by CT scan compared to DXA: A systematic review and meta-analysisClinical Imaging, 2024
- Feasibility study of opportunistic osteoporosis screening on chest CT using a multi-feature fusion DCNN modelArchives of Osteoporosis, 2024
- Artificial intelligence-enhanced opportunistic screening of osteoporosis in CT scan: a scoping ReviewOsteoporosis International, 2024
- CT image-based biomarkers for opportunistic screening of osteoporotic fractures: a systematic review and meta-analysisOsteoporosis International, 2024
- A study on whether deep learning models based on CT images for bone density classification and prediction can be used for opportunistic osteoporosis screeningOsteoporosis International, 2023
- Osteoporosis prediction in lumbar spine X-ray images using the multi-scale weighted fusion contextual transformer networkArtificial Intelligence in Medicine, 2023
- SCOPE 2021: a new scorecard for osteoporosis in EuropeArchives of Osteoporosis, 2021
- Executive Summary of the 2019 ISCD Position Development Conference on Monitoring Treatment, DXA Cross-calibration and Least Significant Change, Spinal Cord Injury, Peri-prosthetic and Orthopedic Bone Health, Transgender Medicine, and PediatricsJournal of Clinical Densitometry, 2019
- OsteoporosisThe Lancet, 2019