A STOCHASTIC FINITE ELEMENT METHOD FOR SIMULATING PATIENT SPECIFIC TRABECULAR BONE
INTRODUCTION: Cancellous bone is a porous, inhomogeneous tissue of moderate stiffness which exhibits anisotropy (directional dependence). Most biomechanical models of cancellous bone incorporate a homogenized continuum mechanics approach which neglects the tissue microstructure. Hence, the method is inherently imprecise and may not be well suited to personalized medicine and other developing clinical techniques. Recent high fidelity trabecular modeling techniques incorporate detailed trabecula microstructure, but are obtainable only with ex-vivo imaging. Thus, existing high fidelity methods have limited applicability in a clinical environment.
RATIONALE: The aim of this research is to create a stochastically accurate high fidelity structural modeling technique to biomechanically represent patient specific trabecular tissue based on existing medical imaging. The method will have applications in clinical precision medicine.
MATERIAL & METHODS: Trabeculae are modeled as beam elements generated via stochastic algorithm to match tissue microstructural distributions. Node positions are assigned randomly with beam connectivity determined by Voronoi diagram. The number of nodes, beam cross sections, and lengths represent the dominant patient factors including bone density, bone adaptation to loading, nutritional status, and the biological response to activity level. The code is general to apply a stochastic approach to all beam properties. Strains were imposed and loads were tracked which permitted extraction of apparent tissue properties.
RESULTS: The stochastic tissue model generation technique produced effective structural properties within published ranges for trabecular bone using inputs which can be based on clinical imaging. The trabecular bone model exhibited a modulus of (198.53 +- 104.5978 MPa). This compares favorably with published values [1,2]. The model bone apparent density was found to be (0.7012 +- 0.2793 g/cm3) which is comparable to established range [1,3]. Additional results will be reported in oral form.
CONCLUSIONS: The proposed finite element technique provides a stochastically accurate structural representation of trabecular tissue and its reaction to applied loads. It incorporate several advantages of high fidelity methods but at lower cost and requiring only clinical imaging. Therefore, the approach may be useful for patient specific musculoskeletal biomechanical models (e.g. osteoporosis, osteoarthritis, joint replacement and implants interface).