Poster Presentation 50th International Society for the Study of the Lumbar Spine Annual Meeting 2024

INTRODUCTION

Computational models are a valuable tool to understand more about spinal loading, but will only provide meaningful results if they represent the spine accurately enough [1]. In-vitro 6-axis testing can assess a spinal segment in all degrees of freedom [2], so may be effective for fine-tuning model material properties and model validation. However, the model and testing must be comparable. A possible source of error is the alignment of the co-ordinate systems, which affects how loading is applied and measured. This sensitivity analysis investigated the effects of co-ordinate system position on model predicted spinal segment stiffness in response to loading, to assess the impact of misalignment.

METHODS

A finite element model of a porcine L2-L3 specimen was developed from MR images to replicate in-vitro 6-axis testing (Fig. 1). The model consisted of an intervertebral disc with adjacent vertebral bodies and longitudinal ligaments, but posterior elements removed. The disc was represented using orthotropic material properties. Experimental stiffness matrix tests were simulated by applying 1.28 mm compression as a preload, then applying set displacements and rotations to the upper vertebral body, with the lower vertebral body fixed. Model predicted forces and moments were assessed against input movements to determine directional spinal segment stiffnesses. The position of the co-ordinate system governing the kinematics and loads was varied by +/- 1 and 2 mm in each direction (anterior-posterior, lateral, superior-inferior), and stiffness changes assessed.

RESULTS

The three principal translation axis stiffnesses were unaffected by shifting the co-ordinate system position. However, the principal rotational stiffnesses were impacted. Stiffness in lateral bending was affected by 5.7% per mm vertical shift in co-ordinate system position, from a baseline of 512 Nmm/degree. Flexion/extension stiffness changed by 10.8% per mm anterior/posterior shift and 6.5% per mm vertical shift, from 290 Nmm/degree. Axial rotation stiffness changed by 4.2% per mm anterior/posterior shift, from 232 Nmm/degree. Differences were also noted in some off-axis stiffnesses. For example, flexion/extension due to a compressive displacement was affected by an anterior/posterior shift in co-ordinate system (stiffness changed by 40.4% per mm shift from a baseline of 894 Nmm/degree). Flexion/extension and axial rotation due to an anterior/posterior displacement were also affected by co-ordinate system position (Fig. 2).

DISCUSSION

Some of the spinal segment directional stiffnesses were unaffected by co-ordinate system position, but several rotational stiffnesses were considerably altered by a millimeter position change. This would suggest that accurate model co-ordinate system position, which impacts how kinematics or loads are applied, is important. It also shows the benefit of direct model validation, where the experimental co-ordinate system is known. A potential limitation of this study was that only a single specimen was assessed, and inter-specimen differences may affect spinal segment behaviour. However, though the amount may vary, the stiffness would likely still be affected. Careful positioning of the co-ordinate system will improve model validation, and allow for more accurate material property calibration, without unintentionally compensating for co-ordinate system malalignment induced stiffness changes, thus aiding more accurate understanding of internal loading or assessments of potential interventions.