A finite element–Komarova framework for CT-based cranial bone remodeling simulation

Bone remodeling is a dynamic biological process in which bone resorption and formation are regulated by mechanical stimuli. Computational modeling of this process requires coupling structural mechanics with biologically driven adaptation laws. This study presents a closed-loop computational framework for simulating adaptive bone remodeling in a bobcat skull ( Lynx rufus ) under bilateral bite loading by integrating micro-CT-based heterogeneous finite element (FE) analysis with the Komarova osteoclast–osteoblast remodeling model. Micro-CT data were converted into heterogeneous elastic properties using Hounsfield Unit (HU)–density and density–elasticity relationships, generating a three-dimensional FE model comprising approximately 12.95 million tetrahedral elements grouped into 500 material bins. Strain energy density (SED) obtained from FE analysis was used to regulate osteoclast and osteoblast activity through the Komarova remodeling equations. Remodeling simulations were performed over 13 iterative cycles, corresponding to approximately 130 simulated days. The results showed a progressive increase in mean Young’s modulus from 4,721 MPa to 6,811 MPa, accompanied by a decrease in maximum SED from 6.93 MPa to 4.83 MPa, indicating a model-predicted structural stiffening response under the prescribed loading condition. In addition, a single-iteration sensitivity analysis across six model, numerical, and loading parameters showed that the remodeling response was relatively insensitive to the individual parameter variations examined, with changes in the mean modulus increment generally below 2% relative to baseline. The proposed framework demonstrated stable large-scale FE–biological coupling and efficient computational performance through a material-bin aggregation strategy and an implicit numerical integration scheme. Overall, this study demonstrates a scalable mechanobiological modeling pipeline for investigating for investigating predicted craniofacial remodeling responses in anatomically complex structures. Key limitations include simplified muscle loading conditions involving only the temporalis muscle and the absence of direct experimental validation; addressing these limitations represents an important direction for future work.