An intramembranous ossification model for the in-silico analysis of bone tissue formation in tooth extraction sites

The accurate modeling of biological processes allows to predict the spatio-temporal behavior of living tissues by computer-aided (in-silico) testing, a useful tool for the development of medical strategies, avoiding the expenses and potential ethical implications of in-vivo experimentation. A model for bone healing in mouth would be useful for selecting proper surgical techniques in dental procedures. In this paper, the formulation and implementation of a model for Intramembranous Ossification is presented aiming to describe the complex process of bone tissue formation in tooth extraction sites. The model consists in a mathematical description of the mechanisms in which different types of cells interact, synthesize and degrade extra-cellular matrices under the influence of biochemical factors. Special attention is given to angiogenesis, oxygen-dependent effects and growth factor-induced apoptosis of fibroblasts. Furthermore, considering the depth-dependent vascularization of mandibular bone and its influence on bone healing, a functional description of the cell distribution on the severed periodontal ligament (PDL) is proposed. The developed model was implemented using the finite element method (FEM) and successfully validated by simulating an animal in-vivo experiment on dogs reported in the literature. A good fit between model outcome and experimental data was obtained with a mean absolute error of 3.04%. The mathematical framework presented here may represent an important tool for the design of future in-vitro and in-vivo tests, as well as a precedent for future in-silico studies on osseointegration and mechanobiology.

💡 Research Summary

The paper presents a comprehensive in‑silico framework for modeling bone tissue formation after tooth extraction, focusing on intramembranous ossification (IO) rather than endochondral pathways. Building on earlier bioregulatory models of long‑bone fracture healing, the authors adapt the mathematical description to the specific anatomy and physiology of the mandibular alveolus. Four cell populations are considered—mesenchymal stem cells (MSCs), fibroblasts, osteoblasts, and endothelial cells—each described by a concentration field (cells per unit volume). Three extracellular matrix (ECM) components (fibrous, bone, and vascular matrices) and two generic growth factors (osteogenic and angiogenic/vascular) complete the system.

The core of the model consists of nine coupled partial differential equations (PDEs). Cell transport combines diffusion, chemotaxis toward growth‑factor gradients, and haptotaxis toward ECM gradients. Proliferation follows a logistic saturation law, while differentiation and apoptosis are represented by linear transfer terms. ECM dynamics include synthesis by specific cell types up to a saturation density and resorption driven by other cells or matrix interactions. Growth‑factor dynamics incorporate diffusion, production by hypoxic cells, natural decay, and consumption by target cells. All equations are nondimensionalized for numerical stability.

Key biological refinements distinguish this work from prior models: (1) removal of chondrogenic variables because IO in the alveolus proceeds without cartilage; (2) explicit modeling of depth‑dependent vascularization, which creates spatially varying oxygen and growth‑factor availability; (3) a novel functional description of cell distribution along the severed periodontal ligament (PDL), reflecting the gradient of cell influx from the surrounding bone. The authors also adopt updated parameter values from recent literature and introduce modifications to improve finite‑element method (FEM) stability.

Implementation uses a two‑dimensional axisymmetric domain representing the extraction socket, discretized with quadrilateral FEM elements. Boundary conditions impose decreasing oxygen and growth‑factor concentrations with increasing depth, mimicking the physiological vascular gradient. Time integration employs an implicit backward‑Euler scheme to ensure robustness against stiff reaction terms.

Model validation is performed against an in‑vivo canine study that measured socket bone volume at multiple time points (0, 7, 14, and 28 days post‑extraction). Simulated bone volume trajectories closely match the experimental data, achieving a mean absolute error of 3.04 %. Sensitivity analysis identifies hypoxia‑dependent apoptosis rates and growth‑factor production coefficients as the most influential parameters on healing outcomes.

The authors discuss strengths such as the integration of multi‑cellular, multi‑signal interactions into a single continuum framework, the ability to capture spatial heterogeneity due to vascular depth, and the potential to inform clinical decision‑making (e.g., selection of graft materials or growth‑factor delivery strategies). Limitations include the two‑dimensional simplification, omission of mechanical feedback between tissue stress and cellular behavior, and reliance on parameter values derived mainly from animal studies, which may require recalibration for human applications.

Future work is proposed to extend the model to three dimensions, incorporate mechanobiological coupling (stress‑dependent differentiation), and calibrate parameters using human clinical datasets. The authors conclude that their intramembranous ossification model provides a valuable predictive tool for dental research, enabling cost‑effective, ethically sound exploration of bone healing strategies and serving as a foundation for subsequent in‑silico studies of osseointegration and tissue engineering.