OBJECTIVES

Predicting implant stability preoperatively remains a challenge. Computed tomography (CT) based finite element (FE) simulations virtually evaluate the mechanical performance of the bone-implant construct. However, translation requires trustworthy simulations based on clinically relevant CT data. The aim of the present study was to evaluate the prediction accuracy of FE models created from cone-beam CT (CBCT) images against experimental results of primary implant stability in human bone specimens.

MATERIAL AND METHODS

Twenty-three dental implants were inserted into bone biopsies extracted from three cadaveric mandibles, and biomechanical testing was performed to determine the load-bearing capacity in a previous study. CBCT-based sample-specific homogenized FE (hFE) models were used to predict ultimate force. The accuracy of the CBCT-based hFE model predictions was compared to the experimental results and to previous μCT-based hFE models.

RESULTS

The ultimate load predicted by the CBCT-based hFE models correlated well with the experimental one (R = 0.66) and was a better estimator than the peri-implant CBCT-based bone density (R = 0.39) or μCT-based bone volume fraction (R = 0.57). Although the results of the two hFE models were strongly correlated (R = 0.91), the μCT-based simulation better predicted the experiments (R = 0.81).

CONCLUSION

By showing that CBCT-based hFE modeling can predict primary stability, this study represents an important step forward toward the clinical translatability of these numerical models as preoperative predictors of primary stability. Nevertheless, several challenges remain to be addressed, such as the lack of an accurate and quantitative way to calibrate CBCT images.