Nonlinear failure prediction in human bone: a clinical approach based on high resolution imaging

The risk of osteoporotic fractures is currently estimated based on an assessment of bone mass as measured by dual-energy X-ray absorptiometry. However, patient-specific finite element (FE) simulations that include information from multiple scales have the potential to allow more accurate prognosis. In the past, FE models of bone were limited either in resolution or to the linearization of the mechanical behavior. Now, nonlinear, high-resolution simulations including the bone microstructure have been made possible by recent advances in simulation methods, computer infrastructure and imaging, allowing the implementation of multiscale modeling schemes. For example, the mechanical loads generated in the musculoskeletal system define the boundary conditions for organ-level, continuum-based FE models, whose nonlinear material properties are derived from microstructural information. Similarly microstructure models include tissue-level information such as the dynamic behavior of collagen by modifying the model’s constitutive law. This multiscale approach to modeling the mechanics of bone allows a more accurate characterization of bone fracture behavior. Furthermore, such models could also include the effects of aging, osteoporosis and drug treatment. Here we present the current state of the art for multiscale modeling and assess its potential to better predict an individual’s risk of fracture in a clinical setting. † Author contributions: DC has written the manuscript. All authors have edited the manuscript. The study was designed by DC and RM.