Computational fluid dynamics simulations for 3D muscle fiber architecture in finite element analysis: Comparisons between computational fluid dynamics and diffusion tensor imaging

Knowledge of skeletal muscle fiber orientations is important for modeling mechanical properties and behavior of muscle tissue. Diffusion tensor imaging (DTI) may be used to define fiber architecture in vivo but can be expensive and time-consuming and thus impractical for biomechanical modeling applications. Muscle tractography using computational fluid dynamics (CFD) has been shown to determine muscle fiber directions for finite element models in which aponeuroses serve as inlet and outlet boundaries. While the technique is simple to implement, it is unclear which flow simulations and constraints achieve fiber architectures similar to DTI and whether FE simulations based on CFD versus DTI fiber directions produce similar results. Here, we implement CFD tractography on the gastrocnemius muscle using a novel boundary condition method that we developed based on specified inflow direction. We compared results from incompressible potential flow and nondimensionalized incompressible Stokes flow. Comparisons were made between flow methods and results from DTI. Mechanical finite element simulations were subsequently performed on muscle models with fiber directions defined by CFD tractography and DTI. Using our boundary condition method, fiber directions modeled from CFD simulations were similar to DTI. Strain distributions from mechanical simulations were similar between Stokes flow and DTI fiber models. This study demonstrates a new method for specifying inlet boundary conditions that generates physiologically reasonable fiber directions in skeletal muscle. Finite element simulations based on this method are similar to those from DTI, illustrating the ability of CFD to determine muscle fiber architecture for modeling purposes when DTI is not available.