The design and application of a diffusion tensor informed finite-element model for exploration of uniaxially prestressed muscle architecture in magnetic resonance imaging

The combination of finite-element models with medical imaging has been a valuable contribution to our understanding of tissue mechanics. In recent years, diffusion tensor imaging has aided in modeling axonal tracts in the brain to measure mechanical stresses related to traumatic brain injuries. Other biological systems and diagnostic techniques can benefit from this approach. Dynamic elastography is a phase contrast imaging technique, where contrast is linked to the mechanical properties (elasticity and viscosity) of the imaged tissue. Mechanical properties are obtained from solving an inverse system based on mechanical wave motion, typically under the assumption that the tissue is homogeneous, isotropic and without initial (pre) stresses or strains. Biological tissues, however, rarely have all three of these properties and the degree to which these assumptions are inaccurate can lead to poor estimates. Muscle typically violates all three major assumptions and requires more refined approaches for elastic moduli estimation. using magnetic resonance-based diffusion tensor (DT) imaging to inform the generation of subject-specific finite-element (FE) models addresses this problem by explicitly accommodating for variations in muscle architecture. This allows for a more robust analysis of prestressed wave motion while compensating for situational geometric changes induced by the loading. The presented work demonstrates a pipeline from DT imaging to FE models and the resulting comparisons with analogous MR elastography experiments. This work will help in developing anisotropic and prestressed relevant inversion algorithms, therefore, improving the accuracy of muscle elastic and viscous moduli estimates.