Spatio-temporal directional filtering for improved inversion of MR elastography images

Manduca, Armando; Lake, David S.; Kruse, Scott A.; Ehman, Richard L.

doi:10.1016/s1361-8415(03)00038-0

Cited by 301 publications

(190 citation statements)

References 8 publications

Supporting

Mentioning

186

Contrasting

Unclassified

Order By: Relevance

“…A directional filter was applied to the displacements to remove backward-propagating (reflected) components based on their position in the Fourier domain [26], [29]. The time delays between each lateral pair of tracking locations were computed via normalized cross-correlation of the tissue velocity signals at each pixel with the same depth.…”

Section: Finite Element Methods Simulationmentioning

confidence: 99%

A multiresolution approach to shear wave image reconstruction

Hollender

Bottenus

Trahey

2015

IEEE Trans. Ultrason., Ferroelect., Freq. Contr.

View full text Add to dashboard Cite

Shear wave imaging techniques build maps of local elasticity estimating the local group velocity of induced mechanical waves. Velocity estimates are formed using the time delay in the motion profile of the medium at two or more points offset from the shear wave source. Because the absolute time-of-flight between any pair of locations scales with the distance between them, there is an inherent trade-off between robustness to time-of-flight errors and lateral spatial resolution based on the number and spacing of the receive points used for each estimate. This work proposes a method of using the time delays measured between all combinations of locations to estimate a noise-robust, high-resolution image. The time-of-flight problem is presented as an overdetermined system of linear equations that can be directly solved with and without spatial regularization terms. Finite element method simulations of acoustic radiation force-induced shear waves are used to illustrate the method, demonstrating superior contrast-to-noise ratio and lateral edge resolution characteristics compared to linear regression of arrival times. This technique may improve shear wave imaging in situations where time-of-flight noise is a limiting factor.

show abstract

Section: Finite Element Methods Simulationmentioning

confidence: 99%

A multiresolution approach to shear wave image reconstruction

Hollender

Bottenus

Trahey

2015

IEEE Trans. Ultrason., Ferroelect., Freq. Contr.

View full text Add to dashboard Cite

show abstract

“…The process of directionally filtering displacement to isolate specific components (directions) of wave propagation has been described and used by others to enhance isotropic material estimates [27]. Direct inversion of MRE displacement fields by fitting to complete wave equations has been used to approximate anisotropic elastic mechanical properties [28,29].…”

Section: Estimation Of Transversely Isotropic Materialsmentioning

confidence: 99%

Mechanical Properties of Viscoelastic Media by Local Frequency Estimation of Divergence-Free Wave Fields

Clayton¹,

Okamoto

Bayly

2013

Journal of Biomechanical Engineering

View full text Add to dashboard Cite

Magnetic resonance elastography (MRE) is an imaging modality with which mechanical properties can be noninvasively measured in living tissue. Magnetic resonance elastography relies on the fact that the elastic shear modulus determines the phase velocity and, hence the wavelength, of shear waves which are visualized by motion-sensitive MR imaging. Local frequency estimation (LFE) has been used to extract the local wavenumber from displacement wave fields recorded by MRE. LFE -based inversion is attractive because it allows material parameters to be estimated without explicitly invoking the equations governing wave propagation, thus obviating the need to numerically compute the Laplacian. Nevertheless, studies using LFE have not explicitly addressed three important issues: (1) tissue viscoelasticity; (2) the effects of longitudinal waves and rigid body motion on estimates of shear modulus; and (3) mechanical anisotropy. In the current study we extend the LFE technique to (1) estimate the (complex) viscoelastic shear modulus in lossy media; (2) eliminate the effects of longitudinal waves and rigid body motion; and (3) determine two distinct shear moduli in anisotropic media. The extended LFE approach is demonstrated by analyzing experimental data from a previouslycharacterized, isotropic, viscoelastic, gelatin phantom and simulated data from a computer model of anisotropic (transversely isotropic) soft material.

show abstract

“…The region behind an inclusion is difficult to reconstruct using the standard methods based on the Helmholtz equation, and typically results in a very noisy reconstruction in that region. To overcome this limitation, special filters are usually applied to eliminate the noise in such a region (see, for example, Manduca et al, 2003). However, the use of such filters could also result in missing inclusions behind other inclusions.…”

Section: Shear Modulus Reconstruction From Mr Measured Datamentioning

confidence: 99%

Shear modulus reconstruction in dynamic elastography: time harmonic case

Park

Maniatty²

2006

Phys. Med. Biol.

102

View full text Add to dashboard Cite

Abstract. This paper presents a direct inversion approach for reconstructing the elastic shear modulus in soft tissue from dynamic measurements of the interior displacement field during time harmonic excitation. The tissue is assumed to obey the equations of nearly incompressible, linear, isotropic elasto-dynamics in harmonic motion. A finite element discretization of the governing equations is used as a basis, and a procedure is outlined to eliminate the need for boundary conditions in the inverse problem. The hydrostatic stress (pressure) is also reconstructed in the process, and the effect of neglecting this term in the governing equations, which is common practice, is considered. The approach does not require iterations and can be performed on subregions of the domain resulting in a computationally efficient method. A sensitivity study is performed to investigate the detectability of abnormal regions of different size and shear modulus contrast from the background. The algorithm is tested on simulated data on a 2D domain, where the data is generated on a very fine mesh to get a near exact solution, then down-sampled to a coarser mesh that is similar to the spatial discretization of actual data, and noise is added. Results showing the effect of the hydrostatic stress term and noise are presented. A reconstruction using MR measured experimental data involving a tissue-mimicking phantom is also shown to demonstrate the algorithm.

show abstract

Spatio-temporal directional filtering for improved inversion of MR elastography images

Cited by 301 publications

References 8 publications

A multiresolution approach to shear wave image reconstruction

A multiresolution approach to shear wave image reconstruction

Mechanical Properties of Viscoelastic Media by Local Frequency Estimation of Divergence-Free Wave Fields

Shear modulus reconstruction in dynamic elastography: time harmonic case

Contact Info

Product

Resources

About