Cerebrospinal Fluid Model Formulation Affects Global and Local Behaviour of the Spinal Cord Submitted to Transverse Traumatic Compression

Abstract

The cerebrospinal fluid (CSF) is a major shock absorber in traumatic spinal cord injuries (SCI). It can be modelled using various fluid formulations, and there is a need to understand their implications on results when performing numerical simulations of SCI. Therefore, four formulations (pressurized volumes – PV, arbitrary lagrangian-eulerian – ALE, smoothed particle hydrodynamics – SPH, and lagrangian – Solid) and two fluid linearity models (Newtonian – N, Mie-Grüneisen – G) were tested in idealised conditions replicating a thoracolumbar burst fracture in a porcine finite element model. The PV formulation proved to be unfit for modelling traumatic SCI, while the solid formulation presented an excessive stiffness of the CSF layer and high computational cost. Both ALE and SPH could be likely candidates, presenting similar results on fragment kinematics, central canal displacement, and anteroposterior deformation of the spinal cord, but dissimilar results in subarachnoid space occlusion, maximal von Mises stress, strain, and volumetric stress. Providing experimental in vivo data on the local dynamics of the spinal will further determine the best formulation to use to replicate the mechanism of injury during traumatic SCI.