Synthetic open cell foams versus a healthy human vertebra: Anisotropy, fluid flow and μ-CT structural studies.

Commercial synthetic open-cell foams are an alternative to human cadaveric bone to simulate in vitro different scenarios of bone infiltration properties. Unfortunately, these artificial foams do not reproduce the anisotropic microstructure of natural bone and, consequently, their suitability in these studies is highly questionable. In order to achieve scaffolds that successfully mimic human bone, microstructural studies of both natural porous media and current synthetic approaches are necessary at different length scales. In this line, the present research was conducted to improve the understanding of local anisotropy in natural vertebral bone and synthetic bone-like porous foams. To attain this objective, small volumes of interest within these materials were reconstructed via micro-computed tomography. The anisotropy of the microstructures was analysed by means of both their main local histomorphometric features and the behaviour of an internal flow computed via computational fluid dynamics. The results showed that the information obtained from each of the micro-volumes of interest could be scaled up to understand not only the macroscopic averaged isotropic and/or anisotropic behaviour of the samples studied, but also to improve the design of macroscopic porous implants better fitting specific local histomorphometric scenarios. The results also clarify the discrepancies in the permeability obtained in the different micro-volumes of interest analysed. STATEMENT OF SIGNIFICANCE: A deep insight comparative study between the porous microstructure of healthy vertebral bone and that of synthetic bone-like open-cell rigid foams used in in vitro permeability studies of bone cement has been performed. The results obtained are of fundamental relevance to computational studies because, in order to achieve convergence values, the computation process should be limited to small computation domains or micro-volumes of interest. This makes the results specific spatial dependent and for this reason computation studies cannot directly capture the macroscopic average behaviour of an anisotropic porous structure such as the one observed in natural bones. The results derived from this study are also important because we have been able to show that the specific spatial information contained in only one healthy vertebra is enough to capture, from a geometric point of view, the same information of "specific surface area vs. porosity" - in other words, the same basic law - that can also be found in other human bones for different patients, even at different biological ages. This is an important finding that, despite the efforts made and the controversies formulated by other authors, should be studied more thoroughly with other bone species and tissues (healthy and/or diseased). Moreover, our results should help to understand that, with the extensive capabilities of current 3D printing technologies, there is an enormous potential in the design of biomimetic porous bone-like scaffolds for bone tissue engineering applications.