Human Mandibular Bone Density Distribution: Image Analysis and Numerical Modeling

Objectives: Bone density distribution affects the short-term and long-term successes of dental treatments. In this study, the bone density distribution in human mandibles is obtained from quantitative analysis on medical images and is also modeled using biomechanics simulations.
Methods: Cone beam computed tomography (CBCT) images from dental clinics. The CBCT scanner is calibrated using our self-made phantoms to convert grayscale to mineral density. The CBCT images of multiple human mandibles are analyzed. Finite element models are used to simulate the effects of normal chewing and biting force on bone density distribution. Strain energy density based bone remodeling algorithm is used in the simulations.
Results: For the same materials in phantoms, there is no significantly difference in its greyscale in different locations inside the region of interests, despite the known drawbacks in CBCT. At the chosen effective energy level, the greyscale in CBCT images has high linearity with attenuation coefficients for the all materials in the phantoms. The averaged trend in mandibular bone density distribution is obtained. It is higher near the root of the tooth and lower away from the tooth, especially near the lingual arc of the jaw. The numerical simulation results show a similar trend in the bone density distribution.
Conclusions: The greyscale in CBCT images from the one CBCT scanner used in this study has linearity and spatial consistency inside the chosen region of interest. The bone density distribution obtained from image analysis has good agreement with that obtained from the bone remodeling simulations based on biomechanics analysis. The results can provide detailed knowledge on dental anatomy and can also build the foundation for future improvements in the dental treatments using image-based biomechanics analysis.