What makes an accurate and reliable subject-specific finite element model? A case study of an elephant femur

• Published in: Journal of the Royal Society interface Vol. 9; no. 67; pp. 351 - 361
• Main Authors: Panagiotopoulou, O., Wilshin, S. D., Rayfield, E. J., Shefelbine, S. J., Hutchinson, J. R.
• Format: Journal Article
• Language: English
• Published: England The Royal Society 07.02.2012
• Subjects: Animals; Biomechanical Phenomena; Biomechanics; Body Size; Bone Mechanics; Cadavers; Convergence; Elephant; Elephantidae; Elephants - anatomy & histology; Female; Femur; Femur - anatomy & histology; Femur - physiology; Finite Element Analysis; Mathematical models; Models, Biological; Simulation; Validation
• ISSN: 1742-5689; 1742-5662
• DOI: 10.1098/rsif.2011.0323

Finite element modelling is well entrenched in comparative vertebrate biomechanics as a tool to assess the mechanical design of skeletal structures and to better comprehend the complex interaction of their form–function relationships. But what makes a reliable subject-specific finite element model? To approach this question, we here present a set of convergence and sensitivity analyses and a validation study as an example, for finite element analysis (FEA) in general, of ways to ensure a reliable model. We detail how choices of element size, type and material properties in FEA influence the results of simulations. We also present an empirical model for estimating heterogeneous material properties throughout an elephant femur (but of broad applicability to FEA). We then use an ex vivo experimental validation test of a cadaveric femur to check our FEA results and find that the heterogeneous model matches the experimental results extremely well, and far better than the homogeneous model. We emphasize how considering heterogeneous material properties in FEA may be critical, so this should become standard practice in comparative FEA studies along with convergence analyses, consideration of element size, type and experimental validation. These steps may be required to obtain accurate models and derive reliable conclusions from them.