Measurement of strain distributions within vertebral body sections by texture correlation

Study Design. A high-resolution strain measurement technique was applied to axially loaded parasagittal sections from thoracic spinal segments. Objectives. To establish a new experimental technique, develop data analysis procedures, characterize intrasample shear strain distributions, and measure intersample variability within a group of morphologically diverse samples. Summary of Background Data. Compression of intact vertebral bodies yields structural stiffness and strength, but not strain patterns within the trabecular bone. Finite element models yield trabecular strains but require uncertain boundary conditions and material properties. Methods. Six spinal segments (T8-T10) were sliced in parasagittal sections 6-mm thick. Axial compression was applied in 25-N increments up to sample failure, then the load was removed. Contact radiographs of the samples were made at each loading level. Strain distributions within the central vertebral body were measured from the contact radiographs by an image correlation procedure. Results. Intrasample shear strain probability distributions were log-normal at all load levels. Shear strains were concentrated directly inferior to the superior endplate and adjacent to the anterior cortex, in regions where fractures are commonly seen clinically. Load removal restored overall sample shape, but measurable residual strains remained. Conclusions. This experimental model is a suitable means of studying low-energy vertebral fractures. The methods of data interpretation are consistent and reliable, and strain patterns correlate with clinical fracture patterns. Quantification of intersample variability provides guidelines for the design of future experiments, and the strain patterns form a basis for validation of finite element models. The results imply that strain uniformity is an important criterion in assessing risk of vertebral failure.