A biomechanical investigation of mandibular molar implants: reproducibility and validity of a finite element analysis model

Background: Three-dimensional finite element analysis (FEA) is effective in analyzing stress distributions around dental implants. However, FEA of living tissue involves many conditions, and the structures and behaviors are complex; thus, it is difficult to ensure the validity of the results. To verify reproducibility and validity, we embedded implants in experimental models and constructed FEA models; implant displacements were compared under various loading conditions. Methods: Implants were embedded in the molar regions of artificial mandibles to fabricate three experimental models. A titanium superstructure was fabricated and three loading points (buccal, central, and lingual) were placed on a first molar. A vertical load of 100 N was applied to each loading point and implant displacements were measured. Next, the experimental models were scanned on micro-computed tomography (CT) and three-dimensional FEA software was used to construct two model types. A model where a contact condition was assumed for the implant and artificial mandible (a contact model) was constructed, as was a model where a fixation condition was assumed (a fixation model). The FEA models were analyzed under similar conditions as the experimental models; implant displacements under loading conditions were compared between the experimental and FEA models. Reproducibility of the models was assessed using the coefficient of variation (CV), and validity was assessed using a correlation coefficient. Results: The CV of implant displacement was 5% to 10% in the experimental and FEA models under loading conditions. Absolute values of implant displacement under loading were smaller in FEA models than the experimental model, but the displacement tendency at each loading site was similar. The correlation coefficient between the experimental and contact models for implant displacement under loading was 0.925 (p < 0.01). The CVs of equivalent stress values in the FEA models were 0.52% to 45.99%. Conclusions: Three-dimensional FEA models were reflective of experimental model displacements and produced highly valid results. Three-dimensional FEA is effective for investigating the behavioral tendencies of implants under loading conditions. However, the validity of the absolute values was low and the reproducibility of the equivalent stresses was inferior; thus, the results should be interpreted with caution.