A Stochastically and Biomechanically Accurate Finite Element Approach for Patient Specific Modeling of Cancellous Bone
INTRODUCTION: Though cancellous bone has high porosity, most published computational models incorporate the use of geometrically homogenized continuum elements (i.e., elements which ignore the porous structure of the tissue). Though requiring only low resolution CT or MRI imaging as an input, required assumptions limit clinical generality. Recently developed high fidelity methods require detailed trabecula microstructure obtainable only with ex-vivo imaging. Thus, they have limited applicability in a clinical setting and may not be useful in personalized medicine. RATIONALE: The purpose of this study is to develop a high fidelity finite element approach useful in personalized medicine to represent the biomechanics of patient specific trabecular tissue. The hypothesis is that a stochastic algorithm driving beam elements can accurately capture the trabecular biomechanics while requiring only conventional CT or MRI imaging. MATERIAL & METHODS: Trabecular bone consists of a three-dimensional network structure mainly composed of rod-shaped and plate-shaped fundamental units named “trabeculae.” The trabeculae are modeled as beam elements created algorithmically and incorporating stochastic distributions that represent the dominant patient factors including bone density, nutritional status, and the biological response to activity level (leading to structural anisotropy). A strain field was imposed, and effective material properties were extracted. RESULTS: The algorithmic approach resulted in an effective structural material properties within published ranges of trabecular bone  while not requiring assumptions such as homogeneity which limit generality. The apparent densities of the current models are also within the published ranges, (1.625 ~ 1.18) g/cm3. CONCLUSIONS: The proposed finite element modeling approach provides a stochastically accurate tissue response, incorporating the advantages of high fidelity models, while requiring only clinical imaging. Thus, it may be useful for patient specific musculo-skeletal biomechanical models. REFRENCES:  Bartel, D.L. Orthopedic Biomechanics, Pearson Education, Inc., 2006.