Osteoporosis is characterized by skeletal fragility where bone quality is deteriorated eventually leading to fractures. In 2017-2018, the prevalence of osteoporosis in adults aged 50 and over was 12.6% and the prevalence of low bone mineral density (BMD), or osteopenia, was 43.1%1. Projections of osteoporosis and osteopenia are estimated to increase to 13.6 million and 57.8 million by 20302. The associated fractures and costs are estimated to increase by 3 million cases and $25.3 billion by 2025, with a cumulative cost of $228 billion3. Due to the continuous increase of the elderly population, obesity, and other associated fracture risk factors (i.e.: diabetes), coupled with non-optimum screening methods for fracture risk prediction, vertebral fractures remain a significant public health burden calling for better solutions and approaches for prevention.

The current generally accepted gold standard method for osteoporosis and fracture risk diagnosis includes dual-energy x-ray absorptiometry (DXA). Although diagnostic criteria have been widely based on areal BMD (g/cm2) from DXA, DXA measurements show site-dependent reliability, with spine measurements often confounded by osteophytes in older individuals4, 5, 6. Additionally, DXA can only account for 2D projections of bone density, and more importantly, cannot estimate bone strength7, 8, 9. These limitations hinder the successful applicability of this technique in the setting of vertebral fracture risk prediction. On the other hand, quantitative computed tomography (QCT) can estimate volumetric BMD (vBMD), accounting for 3D structure, cortical and trabecular bones, and allowing for a more accurate diagnosis of osteoporosis and prediction of fracture risk10, 11, 12. The American College of Radiology (ACR) introduced clinical guidelines to obtain QCT-based vBMD from trabecular bone of the lumbar spine as equivalent measurements for osteoporosis, osteopenia, and normal values, as previously established by the World Health Organization (WHO)13. However, despite being an improvement over DXA, this method is still restricted to providing only a single value that does not consider the structure of the vertebrae and is not capable of estimating the vertebra load bearing capacity.

QCT-based finite element analysis (FEA) significantly enhances fracture risk assessments by precisely considering 3D bone geometry, heterogeneity, and mechanical characteristics, offering a substantial advancement over DXA estimates of aBMD and increasingly attracting clinical interest due to its accurate estimation of bone fracture properties6,8, 14, 15, 16, 17, 18. By implementing QCT-based FEA and considering the previously established guidelines from the ACR, thresholds for vertebral strength for females and males have been previously established for osteoporotic, osteopenic, and normal vertebrae19. However, vertebral FEA models based on QCT have been mostly developed by implementing a calibration phantom which is placed underneath the patient during CT imaging. This process is associated with the quantitative aspect of QCT imaging which includes the conversion of grey values in the CT images (Hounsfield units [HU]) to BMD6,20. Due to significant limitations in using a phantom, including logistics, expenses, and potential technician training, we have developed and validated a phantom-less process that implements the patient's tissues, such as muscle and fat, to convert HU to BMD8. Although the established QCT-based FEA thresholds for fracture prediction marks a significant advancement in the field, it is yet to be determined whether phantom-less methods result in similar thresholds. Therefore, the aims of the current study were twofold: 1) validate vertebral failure load thresholds using a phantom-less approach with previously established thresholds; 2) determine phantom-less and phantom-based vertebral failure load thresholds based on two commonly used material equations and sex. Successful outcomes from this study would further support the use of QCT/FEA as a unique tool in the clinical setting to assess vertebral fracture properties and identify those patients at high risk of fracture. The process could be used prospectively without the need of a calibration phantom and retrospectively on CT scans previously acquired for other purposes.