Requirements for accurate estimation of anisotropic material parameters by magnetic resonance elastography: A computational study

• Published in: Magnetic resonance in medicine 2017-12, Vol.78 (6), p.2360-2372
• Main Authors: Tweten, D.J., Okamoto, R.J., Bayly, P.V.
• Format: Article
• Language: English
• Subjects: Algorithms; Anisotropy; Brain - diagnostic imaging; Computational fluid dynamics; Computer applications; Computer Simulation; Data analysis; Data processing; Elasticity Imaging Techniques; Estimates; Experimental design; Filtration; Finite Element Analysis; Finite element method; heterogeneity; Humans; Image Processing, Computer-Assisted; inversion algorithms; Iron; Low noise; Magnetic resonance; Magnetic Resonance Imaging; Mathematical models; MR elastography; Noise; Noise propagation; Parameter estimation; Propagation; Resonance; Shear modulus; Shear Strength; shear waves; Signal-To-Noise Ratio; Tensile Strength; transversely isotropic material; Traveling waves; Wave propagation
• ISSN: 0740-3194; 1522-2594
• DOI: 10.1002/mrm.26600

Purpose To establish the essential requirements for characterization of a transversely isotropic material by magnetic resonance elastography (MRE). Theory and Methods Three methods for characterizing nearly incompressible, transversely isotropic (ITI) materials were used to analyze data from closed‐form expressions for traveling waves, finite‐element (FE) simulations of waves in homogeneous ITI material, and FE simulations of waves in heterogeneous material. Key properties are the complex shear modulus μ2, shear anisotropy ϕ=μ1/μ2−1, and tensile anisotropy ζ=E1/E2−1. Results Each method provided good estimates of ITI parameters when both slow and fast shear waves with multiple propagation directions were present. No method gave accurate estimates when the displacement field contained only slow shear waves, only fast shear waves, or waves with only a single propagation direction. Methods based on directional filtering are robust to noise and include explicit checks of propagation and polarization. Curl‐based methods led to more accurate estimates in low noise conditions. Parameter estimation in heterogeneous materials is challenging for all methods. Conclusions Multiple shear waves, both slow and fast, with different propagation directions, must be present in the displacement field for accurate parameter estimates in ITI materials. Experimental design and data analysis can ensure that these requirements are met. Magn Reson Med 78:2360–2372, 2017. © 2017 International Society for Magnetic Resonance in Medicine.