Development and validation of a subject-specific finite element model of the functional spinal unit to predict vertebral strength

Finite element models of an isolated vertebral body cannot accurately predict compressive strength of the spinal column because, in life, compressive load is variably distributed across the vertebral body and neural arch. The purpose of this study was to develop and validate a patient-specific finite element model of a functional spinal unit, and then use the model to predict vertebral strength from medical images. A total of 16 cadaveric functional spinal units were scanned and then tested mechanically in bending and compression to generate a vertebral wedge fracture. Before testing, an image processing and finite element analysis framework (SpineVox-Pro), developed previously in MATLAB using ANSYS APDL, was used to generate a subject-specific finite element model with eight-node hexahedral elements. Transversely isotropic linear-elastic material properties were assigned to vertebrae, and simple homogeneous linear-elastic properties were assigned to the intervertebral disc. Forward bending loading conditions were applied to simulate manual handling. Results showed that vertebral strengths measured by experiment were positively correlated with strengths predicted by the functional spinal unit finite element model with von Mises or Drucker-Prager failure criteria (R-2=0.80-0.87), with areal bone mineral density measured by dual-energy X-ray absorptiometry (R-2=0.54) and with volumetric bone mineral density from quantitative computed tomography (R-2=0.79). Large-displacement non-linear analyses on all specimens did not improve predictions. We conclude that subject-specific finite element models of a functional spinal unit have potential to estimate the vertebral strength better than bone mineral density alone.