Studies were performed to determine mechanical properties of fresh human cadaver cervical motion segments, and to develop a computer model of the cervical spine with which to estimate the magnitude and pattern of both muscle contraction forces and segment reaction forces.

In the first study, twenty-five motion segments were mechanically tested, with either fully intact intervertebral ligaments or with only intervertebral discs. These segments were loaded in compression,in three modes of shearing, in three modes of bending, and in torsion. Motions in response to these loads were measured and stiffnesses were computed for the main motions associated with each mode of loading.

When the intact segments were subjected to applied forces (compres- sion or shear), the maximum stiffness was evidenced in compression and the minimum stiffness was evidenced in posterior shear;​ when they were subjected to applied moments (bending or torsion), the maximum stiffness was noted in torsion and the minimum stiffness was noted in flexion. Removal of the posterior elements significantly reduced stiffness in all modes of loading except in posterior shear. No significant differences in stiffness were evident when segments were grouped by vertebral level. Intact segments with greater degrees of disc degeneration generally exhibited larger stiffnesses. Removal of the posterior elements, in general, decreased this effect. Bony union of the posterior articular facets, noted on two segments, markedly increased stiffness in all modes of loading.

In the second study, fourteen volunteer subjects performed a series of exercises involving the cervical musculature. The exercises included maximum voluntary strengths and load-resists, in flexion, extension, lateral bending, and twisting. During the exercises, measurements of surface myoelectric activity were made at the C4 level at eight loca- tions. A computer model, using optimization techniques, was constructed to predict, at the C4 level, for each subject and each exercise, the muscle contraction forces. Myoelectric measurements and predicted forc es were linearly correlated. Correlation coefficients were generally on the order of 0.6 - 0.8 for the anterior zones, 0.5 - 0.85 for the anterolateral zones, 0.35 - 0.55 for the posterolateral zones, and 0.4 - 0.7 for the posterior zones.