Bayesian inference on a microstructural, hyperelastic model of tendon deformation.

Area of Science:

• Biomechanics
• Materials Science
• Computational Modeling

Background:

• Microstructural models offer advantages over phenomenological models for soft tissues by linking parameters to tissue structure.
• Previous attempts to determine these parameters have yielded values with high variability, creating significant uncertainty.
• This uncertainty hinders the reliable application of soft tissue models in areas like artificial tissue design and surgical planning.

Purpose of the Study:

• To derive a novel microstructural hyperelastic model for transversely isotropic soft tissues.
• To apply this model to simulate the mechanical behavior of tendons.
• To incorporate a Bayesian approach to quantify and manage parameter uncertainty within the model.

Main Methods:

• Development of a new microstructural, hyperelastic constitutive model for soft tissues.
• Application of the model to experimental data from tendon mechanical testing.
• Utilisation of an adaptive Markov chain Monte Carlo algorithm for Bayesian parameter estimation.

Main Results:

• The derived model successfully captures the mechanical behavior of tendons.
• The Bayesian approach provided posterior probability distributions for model parameters, quantifying uncertainty.
• The quantified uncertainty was consistent with previously reported parameter ranges.

Conclusions:

• The developed microstructural model effectively addresses parameter uncertainty in soft tissue mechanics.
• The Bayesian framework provides a robust method for quantifying parameter uncertainty in hyperelastic models.
• This approach can serve as a foundation for modeling uncertainty in other soft tissue applications.