Development and validation of a finite element model for lumbar motion segment (L4-L5) / 第二军医大学学报

ABSTRACT

Objective: To develop and validate an accurate three-dimensional geometrical and mechanical finite element(FE) model of the lurabar L4-L5 segment using a new computer-aided designing (CAD) method. Methods: First, a modified "non-seed region segmentation" was done to extract the interest region in the CT image and to obtain a binary image, from which the iso-surface of vertebral body was produced by a discretized marching cubes algorithm. Second, "best cross-section planes" representing the morphologic characteristics of physiological lordosis were used for the initial iso-surface model, forming a "nonregular piecewise subspace". This subspace and the embedded iso-surface model were subsequently transformed by local affine transforms to a "regular subspace", in which a surface mesh of high quality was generated quickly. Finally, a reverse transform procedure was employed to restore the original three-dimensional (3D) image of the lumbar surface mesh of lumbar L4-L5. All coordinate dada of nodal points and message of triangular patches of the surface model were then subjected to ANSYS for the three-dimensional FE mesh construction. An accurate 3D non-linear FE model of lumbar motion segment (L4-L5) was developed and validated against published data. Results: The constructe d FE model of lumbar L4-L5 consisted of 94 794 solid elements, 1 196 link elements, 1 170 shell elements, 768 target elements and 464 contact elements, and included geometrical, material and contact non-linearities. The predicted results of lumbar L4-L5 segment were closely correlated with published results of experimental biomechanics in terms of axial displacement, segment rotation and intradiscal pressure under similar load condition. Conclusion: Based on advanced algorithm, this constructed surface model of L4-L5 segment is capable to perform whole digitalized binary image extraction and reconstruction of the lumbar surface with excellent simulation results.