“…Different modelling approaches will be likely necessary to further improve the results. Among them, in the authors' perspective, some promising ones are multiscale models to determine bone properties (Vaughan et al, 2012), and imageprocessing algorithms that de-blur CT images (Treece et al, 2012;Pakdel et al, 2014) or identify bone anisotropy (Larsson et al, 2014). Nonetheless, this shortcoming of our results does not impair the application of our models to clinical studies, as already done by Keyak et al (2011) with a model based on Keyak et al (2005) that did not reach a X¼Y relationship.…”