Influence of key modeling assumptions on the finite element prediction of dental implant primary stability.

The clinical success of a dental implant is dependent on its post-operative primary stability. Finite element (FE) modeling of the bone-implant construct stability has gained interest, but the impact of modeling assumptions on outcomes remains uncertain. This sensitivity study examines the effect of three key parameters - bone homogeneity, bone-implant contact definition and implant properties - on the load-bearing capacity of the bone-implant construct. A validated FE model served as a reference, replicating the experimental loading of dental implants inserted into 30 human jawbone biopsies from a previous study. This base model incorporated heterogeneous bone, contact with friction at the bone-implant interface, and a rigid implant. Model variations were generated by independently changing each parameter, the alternative modeling choices being homogeneous bone, a tied interface and an elastic implant, respectively. Model accuracy was evaluated based on correlation and root mean square error with the experimental ultimate force. Homogeneous bone properties led to an underestimation of the ultimate load, while a tied interface overestimated it. Modeling homogeneous bone or a tied interface resulted in lower correlations with experimental ultimate force (R2 = 0.61 and 0.70, respectively), compared to the base model (R2 = 0.81). Implant material (elastic vs. rigid) impacted the initial stiffness but not the load-bearing capacity. Employing heterogeneous bone properties and defining the interface as a contact with friction is recommended to accurately model primary stability of bone-implant constructs. In particular, an overestimation of the primary stability could lead to suboptimal implant design, placement or biased decisions.