Measurement of intervertebral movements is essential in the clinical diagnosis and assessment of low back pain and spinal disorders such as instability. It is possible to measure gross movements of the entire lumbar spine using markers or sensors attached to the skin. But such techniques are not accurate enough to provide information about intervertebral movement as the magnitude of the movement is similar to that of the error due to soft tissue deformation underlying the sensors. Radiographic methods are able to provide accurate data, but this involves identification and superimposition of vertebral images which is a time consuming and technically demanding process. This thesis attempted to address the above limitations by two related studies. The purpose of the first study was to examine the feasibility of an inverse kinematic algorithm in predicting intervertebral movements using information derived from skin-mounted sensors. The second study involved the development of an automatic method of identifying vertebral images in radiographs and computing their movements. This would significantly reduce data processing time and increase the attraction of the radiographic method as a clinical tool for measuring intervertebral movements.; In the first study, an inverse kinematic model was employed to determine the optimum intervertebral joint configuration for a given forward-bending posture of the human spine. An optimization equation with physiological constraints was employed to determine the intervertebral joint configuration. Experimental validation was performed using lateral radiographs of the lumbosacral spines of twenty-two subjects (9 men and 13 women, 40 +/- 14 years old). The model was found to be valid for predicting the intervertebral rotations of the lumbar spine segments but not intervertebral translations. The differences between the measured and predicted values of intervertebral rotations were generally small (less than 1.6 degrees). Pearson product-moment correlations were found to be high, ranging from 0.83 to 0.97, for prediction of intervertebral rotation, but poor for intervertebral translation (R = 0.08 to 0.67). The inverse kinematic model can be clinically useful for predicting intervertebral rotation when X-ray or invasive measurements are undesirable. It is also useful in biomechanical modeling, which requires accurate kinematic information as model input data.; Knowledge of the intervertebral translations of the spine is essential in the clinical assessment of some clinical disorders such as instability, spondylolysis or spondylolisthesis. Unfortunately, such assessment could not reliably performed using the inverse kinematic method but only by radiographic measurement. In the second part of this study, the precision and accuracy of a new automatic method to determine intervertebral movements were examined. Active contour was employed for segmentation of vertebral body image, providing a rapid and accurate measurement of vertebral shape using Fourier descriptors. A Genetic Algorithm was then utilized to determine the displacements of the vertebral bodies. Lateral radiographs of the lumbosacral spines of twenty-two healthy male volunteers (21 +/- 1 years old) were taken in full flexion and extension. The vertebral body image was fitted with a quadrangle and its corners to be digitised. This allowed the intervertebral movement to be determined manually. The mean differences in the angles determined by the manual and automatic method were less than 1.4 degrees; whereas the mean differences in posterior-anterior and superior-inferior translations less than 1.2 mm and 0.8 mm respectively. The correlation of vertebral body outline as determined by the automatic method in the flexion and extension films was high, with R values ranging from 0.994 to 0.997. This indicates that no image distortion or out-of-plane movements occur. The root mean square error of data among five repeated measurements were less than
展开▼
