Model developments for in silico studies of the lumbar spine biomechanics

Tesis doctoral de Jerome Bernard Noailly

This phd thesis investigated the use of finite element modelling to study lumbar spine biomechanics for clinical assessment. Bibliographic studies reported in the first chapter showed clear functional relations between external forces and lumbar spine tissue structures and shapes. Clinical research revealed that independently of its origin, low back pain may be worsened by altered tissue mechanical environments. Experimental measurements alone cannot truly describe the load distributions between the different lumbar spine tissues. Thus, finite element models have been used in the past. But model reliability in predicting local tissue loadings is still not manifest and has been explored in this thesis as described in the following chapters. in chapter 2, a l3-l5 lumbar spine bi-segment model was built. An initial model was completed to include the vertebral cortex, a full definition of the facet joints, the cartilage endplates, and an improved description of the annulus fibre-reinforced structure. Simplified load-cases used for in vitro studies were simulated to calculate stress and strain energy distributions. Predictions within the l3-l5 lumbar spine bi-segment model could be interpreted in terms of functional load distributions related to known tissue structures, but the overall l3-l5 bi- segment model geometry needed further update. thus, in chapter 3, a geometrically accurate l3-l5 lumbar spine bi-segment model was created. The new model included corrected l3 and l5 body shapes and dimensions, corrected disc heights and nucleus placements, corrected posterior bone shapes, dimensions, and orientations, and corrected ligament distributions. The new and old geometries were biomechanically compared. Results showed that the relative roles of modelled tissues greatly depend on the geometry. Predicted load distributions were generally more physiological in the new model. However, new and old models could both reproduce experimental ranges of motion, meaning that their validation should take into account local load transfers. chapter 4 focuses on the variability of the annulus collagen criss-cross angles. Four bi-segment models with literature-based annulus fibre organizations were created and compared under diverse loads. Moreover, an annulus stabilization parameter was proposed by analogy to a thick walled pipe. Model biomechanics greatly depended on the annulus fibre organization, but annulus stabilization parameter was often contradictory with the predicted stresses and strains. Spine geometry and annulus fibrous organization were hypothesized to be linked together. Adapted annulus collagen networks may be numerically determined, but annulus modelling should be based on mechano-biological relationships. in chapter 5, a case-study of a novel artificial disc design coupled with the l3-l5 lumbar spine model is presented. Bi-segment models with and without implant were compared under load- or displacement-controlled rotations, with or without body-weight like load. Prosthesis stiffness generally altered the load distributions and displacement-controlled rotations led to strong adjacent level effects. Including body weight-like loads seemed to give more realistic results. Although the novel disc substitute is too stiff, it is more promising than other existing commercial devices. in this thesis, six new osteoligamentous lumbar spine bi-segment finite element models were created. simulations showed that reliable use of lumbar spine finite element models requires precise descriptions of local tissue loading and response. Local predictions were qualitatively mainly limited by a lack of knowledge about soft tissue structural organisations, constitutive equations, and boundary conditions. However, models can be used as in silico laboratories to overcome such limitations. A hierarchical procedure for the development of qualitatively reliable lumbar spine finite element models was proposed based on available numerical and experimental inputs.

 

Datos académicos de la tesis doctoral «Model developments for in silico studies of the lumbar spine biomechanics«

  • Título de la tesis:  Model developments for in silico studies of the lumbar spine biomechanics
  • Autor:  Jerome Bernard Noailly
  • Universidad:  Politécnica de catalunya
  • Fecha de lectura de la tesis:  22/06/2009

 

Dirección y tribunal

  • Director de la tesis
    • Damien Lacroix
  • Tribunal
    • Presidente del tribunal: Francisco javier Gil mur
    • luigi Ambrosio (vocal)
    • hans-joachim Wilke (vocal)
    • Manuel Doblaré castellano (vocal)

 

Deja un comentario

Tu dirección de correo electrónico no será publicada. Los campos obligatorios están marcados con *

Scroll al inicio