A Computationally Efficient and Accurate Lumbar Spine Model

The design of spine implants requires a good understanding of spine kinematics and loading conditions. Realistic simulation of each functional spinal unit (FSU) requires capturing complicated contact and deformation of biological tissues in a computationally efficient manner. Specifically, the complexities include contacts in intervertebral and facet joints, restraints of spine ligaments, as well as realistic material properties of soft tissues. The variation in the stiffness among different FSUs is often neglected in spine modeling, which might be crucial for spine function. A hybrid approach for lumbar spine modeling was established that combined motion capture experiments, kinematic spine modeling and detailed finite element modeling. Motion capture data during flexion was collected and used to drive the spine model. For computational efficiency each FSU was modeled as an intervertebral connector (joint) element with different elastic behavior at each level. The connector behavior was calibrated using experimental data on the whole lumbar spinal motion (Wong et al. 2006) and cadaveric moment-rotation relationship of L45 (Heurer et al. 2007). Then the predicted stiffness for L23 was used to calibrate the material properties of a detailed FEM of L23.