An increasing number of patients suffer from hip joint degeneration. One possible treatment is total hip replacement surgery, during which the damaged joint is replaced by an artificial hip implant. A common complication that arises after the implantation is stress shielding, characterized by reduced stresses in the remaining thigh bone. Caused by too rigid implants, the reduced stresses may result in bone resorption, which in turn increases the risk of implant loosening. A countermeasure found in literature are bone implants with internal lattice structures, e.g., described in [1]. Inspired by porous bone, the lattice structure offers additional degrees of freedom for design optimization. Recent advances in additive manufacturing allow to realize the thought-up lattice designs. As a secondary benefit, the lattice structures facilitate bone ingrowth. To predict the performance of such lattice-based implants and ultimately optimize their design, corresponding geometric models are required. In contrast to the most common modelling approaches that result in lattice structures which do not conform with the implant’s outer shape, in this work we use a boundary-conforming spline composition approach [2, 3]. This approach allows for both homogeneous and locally varying lattice structures. In particular, local variations within the lattice structures are possible, while still ensuring conformity between individual cells, if parametrized unit cells in combination with continuous parameter fields are used. This permits to functionally grade the lattice structures. This presentation showcases the construction of spline-based geometry models, starting from bone scans and solid implant designs. The resulting geometry models can represent both the natural and implanted configurations, where the implants can be solid or comprise lattice structures. These smooth geometry representations directly enable isogeometric analysis. Moreover, isogeometric analysis facilitates the use of gradient-based optimization methods, which can be employed to improve the implant design.