Stochastic multiscale modelling of cortical bone elasticity based on high-resolution imaging.

Area of Science:

• Biomechanics
• Materials Science
• Computational Modeling

Background:

• Accurate bone quality assessment needs methods to predict mechanical properties from microstructure.
• X-ray-based methods for bone microstructure can yield uncertain data, especially in vivo.
• Deterministic models struggle with input uncertainties, necessitating new approaches.

Purpose of the Study:

• To develop a novel stochastic multiscale model for estimating bone elastic properties.
• To incorporate uncertainties in bone composition (collagen, mineral, water) into the model.
• To provide reliable statistical information on bone mechanical properties despite input data limitations.

Main Methods:

• Developed a stochastic multiscale model based on continuum micromechanics.
• Treated bone component volume fractions as random variables using the maximum entropy principle.
• Utilized synchrotron radiation micro-computed tomography data from a human femoral neck sample.

Main Results:

• Computed effective elastic properties of cortical bone tissue.
• The stochastic model yielded reliable statistical data (mean values, confidence intervals) for bone elastic properties.
• Identified a simpler "nominal model" that captures the stochastic model's key features.

Conclusions:

• Stochastic multiscale modeling offers a reliable approach to predict bone mechanical properties with uncertain input data.
• This method enhances the accuracy of bone quality assessment at the tissue scale.
• The findings are significant for understanding bone mechanics and developing predictive tools.