Etude biomecanique de la degenerescence du disque intervertebral a l'aide d'un modele elements finis poroelastique.

摘要

The intervertebral discs impact the flexibility and mobility of the spine and play an important role in transmitting loads through the spine. Disc wear and degeneration occur as a result of age and environmental factors while these changes affect the main function of the disc. The degeneration is generally associated with low back pain, mainly in the lumbar region, where the spine carries heavy loads. Experimental studies were realized on cadaveric functional units to investigate this pathology and to understand its effect on spinal mobility. As a complement, poroelastic finite element (FE) models have been developed and used to represent the biphasic behaviour of the disc. The inclusion of this poroelasticity is essential in the representation of the degeneration process. However, to date such models were not really used to study the degenerative pathology by itself. The purpose of this project was to develop and validate a poroelastic parametric FE model of the intervetebral disc and then identify the significant biomechanical parameters affecting the healthy and degenerated disc behaviour.;The second objective was to determine discal properties significantly affecting the biomechanical behaviour of healthy (GR.I) and degenerated disc (GR.IV) models. Disc height (H), fiber proportions (%F), drained Young's modulus (Ea, En) and initial permeability (ka, kn) of both annulus (a) and nucleus (n) were analyzed using an experimental design. Modalities of these parameters were set as +/-40% above and below mean values used for healthy (GR.I) and degenerated (GR.IV) discs. A total of 16 simulations were performed for every combination of disc grades and loading cases (compression, lateral bending, flexion and extension). To determine the significant influence (p-value 0,05) on the biomechanical behaviour of the disc model, mobility (ROM and strain rate after 1, 5, 45, 125 and 245 minutes of creep deformation) and load transfer (PP and effective stress profiles in discal region at 1, 5, 45, 125 and 245 minutes of loading application) were analyzed. The statistical analysis showed that disc height had a significant influence (p 0.05) on the overall biomechanical behavior for both healthy and degenerated disc models during the entire loading history. For all loading cases, the annulus' Young modulus significantly affected SE in the annulus zone for both disc grades, but was also significant in the nucleus zone for the degenerated discs with further creep response. Permeability had a significant influence on PP stress for both disc grades, but this effect occurred earlier in the degenerated discs when compared to healthy discs.;This study includes some limitations. First, disc height, which is considerably altered by the process of disc degeneration, was the only modified geometrical parameter used to represent the degenerated disc geometry. However, the use of a generic geometry does not strictly correspond to the personalize disc degeneration observed in vivo. Nevertheless, the FE-predicted creep curves agreed well with those from the literature. The ROM FE-predicted in lateral bending, flexion and extension was inferior to published ROM. However, the trends were similar, and the ROM decreased with disc degeneration. Osmotic pressure, which offers a supplementary resistance to fluid movement, was not considered in the model. This probably affected the response of the healthy disc, but this omission remained less important in the degenerated disc behavior, which behaved more like a solid material.;The work carried out in the framework of this project demonstrated the difference between mobility and load-sharing for healthy and degenerated disc models. The developed modeling approach allowed the representation of disc grades by altering mechanical and geometrical parameters associated with the degeneration process. Healthy disc behavior was mainly carried by nucleic fluid, whereas degenerated disc behavior was mainly carried by the solid phase. This modeling work distinguishes itself from other published models as, for the first time, the parameters that affect biomechanical behaviour of healthy and degenerated disc were identified. Further studies should be performed to include personalized disc properties with the help of quantitative imaging techniques. Futhermore, the model should include posterior elements and should be extended to include the complete lumbar segment in order to perform extensive studies on the degenerated disc response. More specifically, such as studies may explore the effects of degenerated discs on the adjacent levels or the degenerative impact on the facet joint and ligaments. Moreover, the model should be used to study the impact of posture and dynamic solicitation on the biomechanical behaviour of healthy and degenerated discs.;The first objective of this project was to develop and validate a poroelastic parametric FE model, which integrates a generic representation of the disc structure, endplates and vertebral bodies. The simplified geometry was generated using published parametric equations and data found in the literature. Disc height and poroelastic properties of both healthy (Thomson grade I) and degenerated (Thompson grades III and IV) discs were also taken from published data. FE models were validated using published experiments exploring creep. Ranges of motion (ROM) in lateral bending, flexion and extension were compared with those from published in vitro experiments. Pore pressure (PP), effective stress (SE) and total stress (ST) profiles were analyzed as a function of time following the application of load along discal region profiles for each disc grade. The relative contribution of S E and PP was then analyzed as function of time and in the mid-sagittal region. To do so, a compressive stress of 0,35 MPa was applied for every loading case and a moment of 5Nm was added. Simulations of grades I, III and IV disc models using the compressive stress alone agreed well with available published experimental creep data. However, ranges of motion obtained from bending moments were lower than published experimental values. As compared with healthy disc models, stress profiles were mainly concentrated in the annulus region for degenerated disc models (principally in the compressed zone). The PP was dissipated as consolidation occurs, at a higher rate for highly degenerated discs (GR.IV). Then, as the fluid was expulsed, the solid matrix took up extra stresses. For healthy discs, the majority of stress was carried out by the fluid for the entire loading time. Conversely, the majority of stress was undertaken by the solid matrix at the end of the loading application for the degenerated discs.