Experimental and Numerical Investigation of the Bone-Screw in Axial Pull-Out Test

Pedicle screw fixation is a widely used technique for spinal stabilization, especially in patients with osteoporosis or other conditions that affect bone quality. The mechanical stability of pedicle screws is influenced by bone density and structural properties, which can be evaluated using both experimental and numerical methods. The purpose of this study was to evaluate the pull-out strength of pedicle screws in vertebral trabecular bone using both experimental testing and finite element analysis (FEA).In the experimental phase, pull-out tests were performed on normal, osteopenic, and osteoporotic vertebral specimens, followed by micro-CT imaging to assess bone microarchitecture, including bone volume fraction (BV/TV) and trabecular thickness. Since direct integration of micro-CT data into FEA was not feasible, BV/TV values were used to estimate apparent bone density and assign material properties in the FE model. Finite element simulations were then performed using a homogeneous material approach for comparison with experimental pull-out forces (PPF). The results showed a strong correlation between FE and experimental pull-out forces, with an average discrepancy of approximately 10%. However, variations between FE and experimental PPF were observed in different bone quality groups, especially in osteoporotic bone. In addition, a mesh sensitivity analysis was performed to determine the optimal mesh size for accurately capturing screw-bone interactions.Despite certain limitations, including material homogeneity assumptions and segmentation challenges in micro-CT data, this study demonstrates the feasibility of using FEA to predict pedicle screw pull-out strength. The results improve computational modeling techniques for spinal fixation and highlight the importance of bone quality in surgical outcomes. Future studies should focus on refining FE modeling approaches, incorporating patient-specific bone structures, and integrating advanced material models to improve prediction accuracy.