The purpose of this study was to develop a technique for estimating the three force and three moment components acting on the human spine. The technique is based on measuring the three force and three moment components acting on a clinically used external fixation device and modelling the relationship between the forces and moments acting on the fixation device and the forces and moments transmitted along the spine. In order to describe the load sharing between the fixation device and the spine, a mechanical and a statistical model were developed. The mechanical model consisted of a model describing the load- deformation behavior of the fixation device, and one describing the load- deformation behavior of the spinal segments. The statistical model was derived from load sharing relationships between the fixation device and spinal segments which were obtained from in vitro experiments.

Mechanical tests were performed to investigate the performance of the part of the mechanical model which described the load-deformation behavior of the fixation device. The results of these tests indicated that the load transmitted along the fixation device could be estimated within an accuracy of 6.3% for the forces and 12.9% for the moments. The estimation of the deformation of the fixation device showed errors in the order of magnitude of 30% which may be decreased by modelling the relative movement at the interconnections.

In vitro tests were performed to investigate the performance of the mechanical and the statistical model. The results of these tests indicated that the models were most sensitive to changes in the geometry and stiffness of the spinal segments and the stiffness of the bone-metal inter -faces. The individual variables could be controlled in the mechanical model, however, they could not be controlled in the statistical model. These variables influenced the results in a subtractive and additive way.

The results of this study showed that the mechanical model combined with the force and moment measuring technique, could be a powerful tool when used to estimate the three-dimensional spinal load situation. Before this technique can be clinically useful, a detailed assessment of the influence of biological and anatomical variations in the spinal segments on the performance of the mechanical model is required.