Optimization of digital volume correlation computation in SR-microCT images of trabecular bone and bone-biomaterial systems.

ABSTRACT

A micromechanical characterization of biomaterials for bone tissue engineering is essential to understand the quality of the newly regenerated bone, enabling the improvement of tissue regeneration strategies. A combination of microcomputed tomography in conjunction with in situ mechanical testing and digital volume correlation (DVC) has become a powerful technique to investigate the internal deformation of bone structure at a range of dimensional scales. However, in order to obtain accurate three-dimensional strain measurement at tissue level, high-resolution images must be acquired, and displacement/strain measurement uncertainties evaluated. The aim of this study was to optimize imaging parameters, image postprocessing and DVC settings to enhance computation based on 'zero-strain' repeated high-resolution synchrotron microCT scans of trabecular bone and bone-biomaterial systems. Low exposures to SR X-ray radiation were required to minimize irradiation-induced tissue damage, resulting in the need of advanced three-dimensional filters on the reconstructed images to reduce DVC-measured strain errors. Furthermore, the computation of strain values only in the hard phase (i.e. bone, biomaterial) allowed the exclusion of large artefacts localized in the bone marrow. This study demonstrated the suitability of a local DVC approach based on synchrotron microCT images to investigate the micromechanics of trabecular bone and bone-biomaterial composites at tissue level with a standard deviation of the errors in the region of 100 microstrain after a thorough optimization of DVC computation. LAY DESCRIPTION Understanding the quality of newly regenerated bone after implantation of novel biomaterials is essential to improve bone tissue engineering strategies and formulation of biomaterials. The relationship between microstructure and mechanics of bone has been previously addressed combining microcomputed tomography with in situ mechanical testing. The addition of an image-based experimental technique such as digital volume correlation (DVC) allows to characterize the deformation of materials in a three-dimensional manner. However, in order to obtain accurate information at the micro-scale, high-resolution images, obtained for example by using synchrotron radiation microcomputed tomography, as well as optimization of the DVC computation are needed. This study presents the effect of different imaging parameters, image postprocessing and DVC settings for as accurate investigation of trabecular bone structure and bone-biomaterial interfaces. The results showed that when appropriate image postprocessing and DVC settings are used DVC computation results in very low strain errors. This is of vital importance for a correct understanding of the deformation in bone-biomaterial systems and the ability of such biomaterials in producing new bone comparable with the native tissue they are meant to replace.