In vivo measurement of vertebral deformation using digital tomosynthesis based digital volume correlation

Vertebral fractures are the most common type of osteoporotic fracture and associated with significant complications. Timely intervention is important to prevent vertebral fractures, however the current standard for assessing osteoporosis (bone mineral density) is not fully accurate for identifying at-risk individuals. Inspired by a laboratory technique combining microcomputed tomography with mechanical loading for mechanical assessment of extracted bone structures, digital tomosynthesis-based digital volume correlation (DTS-DVC) uses supine and standing DTS images of patients in combination with DVC. The current study evaluated in vivo precision errors, and the utility of DTS-DVC in identifying mechanically compromised vertebrae. Seven patients with vertebral fracture (Fx) and twelve without (NFx) were DTS-imaged, and endplate-to-endplate displacement, stiffness, compliance, and endplate distribution statistics were calculated using supine reference images and images acquired in supine, standing, standing while holding added weight. The in vivo measurement error of DTS-DVC metrics and the extent to which DTS-DVC can measure differences in vertebrae due to loading and presence of vertebral deformity (vertebral fracture) were examined. Total measurement error was low (0.017-0.019 mm), and all measured parameters changed with loading (p < 0.0001 to p < 0.05). Endplate-to-endplate displacement and displacement heterogeneity were significantly higher in fractured vs adjacent intact vertebrae. There were large differences in DVC variables between intact L1 vertebrae of Fx and NFx groups; however, these were not statistically demonstrable. Collectively, results support the in vivo feasibility of DTS-DVC and warrant further investigation. A biomechanics-based assessment of vertebral bone quality is expected to improve our understanding and clinical assessment of vertebral fracture risk.