註釋 The work performed in the present Master's Thesis entailed the assessment of two problems of clinical interest through two separate studies based on finite element analysis (FEA). The first study evaluated the biomechanical behaviour of dental prostheses supported by implants with different diameters. In this context, narrow-diameter implants (NDIs) have been traditionally identified with higher rates of failure in comparison with standard-diameter implants (SDIs) and widediameter implants (WDIs) since they generate a more unfavourable stress distribution. However, it is well known that the load sharing effect associated with prostheses supported by multiple implants (also called splinted prostheses) involves mechanical benefits. Therefore, this study was conducted to evaluate by means of FEA whether the risks linked to NDIs could be mitigated by the mechanical advantages related to the splinting concept. For that purpose, a 3D model of a real maxilla was reconstructed from images of computed tomography (CT) and different implants (NDIs, SDIs and WDIs) and prostheses were created by using computer-aided design (CAD) tools. Biting forces were simulated on the prostheses of three different rehabilitation solutions: single-implant restoration, 3-unit bridge and All-on-4 treatment. The mechanical benefits of the splinting concept were verified: in comparison with unsplinted NDIs, splinted NDIs supporting the 3-unit bridge implied average reductions in terms of volume overloaded under compression and tension, in that order, by 32% and 64% in cancellous bone and 76% and 73% in cortical bone. However, splinted NDIs supporting the typical full-arch of the All-on-4 treatment resulted in the highest risk of overloading found in the study due to the increase of the compressive stress levels generated around the tilted implant when loading the cantilevered molar. On the other hand, the second study was carried out to compare, in terms of microarchitecture and permeability, a human L3 vertebra and two commercial synthetic foams. These foams are used in in vitro permeability tests for vertebroplasty studies under the assumption that their microarchitecture is similar to that of real vertebrae. The methodology was based on the reconstruction of 3D models from micro-computed tomography (μCT) images and the computation of flow dynamic simulations in different regions of the models. On average, vertebral bone revealed a more intricate and compact microstructure than the one of the foams. These findings agree with the lower values of permeability found in the vertebra when compared with the foams. Moreover, foams showed an isotropic architecture whereas the nature of the vertebra was clearly anisotropic. Finally, among the representative histomorphometric indices for porous materials, the index of specific surface area (BS/TV) was the one that better captured the relationship between the permeability and the internal microarchitectural features of the samples analysed. 