Finite element simulation of the healthy and degenerated lumbar spine: interplay between muscle activity and intervertebral disc multiphysics
- Autores: Themis Toumanidou
- Directores de la Tesis: Jerome Bernard Noailly (dir. tes.), Josep Maria Font Llagunes (tut. tes.)
- Lectura: En la Universitat Politècnica de Catalunya (UPC) ( España ) en 2016
- Idioma: español
- Tribunal Calificador de la Tesis: Ferrán Pellisé Urquiza (presid.), Estefanía Peña Baquedano (secret.), Fabio Galbusera (voc.)
- Materias:
- Enlaces
- Tesis en acceso abierto en: TDX
- Resumen
The human spine provides mechanical support to the trunk while it protects the spinal cord and nerves from the external loads transferred during daily activities. Such loads are largely controlled by the spine muscles and influence the biophysical regulation of the intervertebral discs (IVD). Numerical models have been important tools for the translation of the external forces into internal loads that otherwise cannot be easily measured directly. This PhD thesis used the predictive ability of constitutive equations to reflect the mechanical properties of the lumbar IVD and muscles and explore the IVD-muscle interplay on the healthy and degenerated spine.
A review of the state-of-the-art reported for the estimation of spine loads was performed, and the Hill¿s mus cle model and the poro-hyperelastic formulations used for IVD modeling were particularly detailed. A new constitutive equation assembly was proposed involving one active parameter controlled via strain-based criteria, and four passive parameters. For the latters, literature-based values were initially defined, and a parametric study was designed for the active parameter by proposing stretch-related activation thresholds. An optimization scheme was then developed to define a full set of calibrated values per fascicle using force estimations from a reported rigid body model based on measured kinematics of the vertebrae.
To test the robustness of the method, a generic L3-S1 finite element (FE) model was developed that included 46 muscle fascicles and all passive issues. Simulation of forward flexion showed that the predicted muscle forces increased in caudal direction. The intradiscal pressure (IDP) predictions correlated with previous in vivo measurements showing the ability of the model to capture realistic internal loads. To simulate standing, the gravity loads were defined by considering the heterogeneous distribution of body volumes along the trunk. This simulation was also coupled to a previous 8-hour free IVD swelling to mimic the overnight disc hydration. Disc swelling led to muscle activation and force distributions that seemed particularly appropriate to counterbalance the gravity loads, pointing out the likely existence of a functional balance between stretch-induced muscle activation and IVD multiphysics.
A geometrical extension was then performed to incorporate all relevant tissues of the full lumbar spine including in total 96 fascicles. The effect of previous rest (PR) and muscle presence (MS) on internal loads was explored in standing and lying.
Muscle force predictions in standing showed that with PR, the total loads transferred were altered from compressive to tensile. Overnight, the computed IDP increase reproduced previous in vivo data. Both PR and MS affected the vertebrae motion particularly between L1-L2. When degenerated discs properties were used, a general IDP decrease and up to 14 times higher activation was predicted in standing with PR.At last, the previous workflow was repeated using a patient L1-S1 FE model with patient-specific (P-SP) and condition-depended material properties. In standing, asymmetric fascicle activation with increased shortening at the left side and lateral bending was predicted. The decreased swelling capacity of the degenerated discs was associated to an increased muscle activation needed to balance the gravity loads that tended to flex forward the trunk. Comparisons of the IDP results in both models with healthy discs showed that introducing P-SP geometries gave better correlations with in vivo data. Given the difficulties to evaluate the predicted muscle forces experimentally, such outcome further contributed to the validation of the method. Despite its limitations, this approach allowed to explicitly and rationally explore the interactions between muscle function and passive tissue biomechanics in the lumbar spine. The information provided could help clinical decision for patients whom source of back pain is unclear
