Implementing Magnetic Resonance Elastography in the MAGNUS for Evaluation of High-Performance Gradient Coil Effects

Introduction: Magnetic resonance elastography (MRE) is a method capable of non-invasively quantifying tissue mechanical properties[1] and has recently developed into a tool for the study of the brain[2]. Mechanical characterization of brain tissue through MRE has been used to study neurodegeneration[3], brain injury[4], development[5], and cognition[6].
Development of MRE methods for optimal performance have recently focused on image acquisition techniques[7], motion encoding schemas[8], and inversion algorithms[9], all of which focus on improving one or two of the following categories: high spatial resolution, good signal to noise ratios (SNR), short scan time, high localized accuracy, and good motion encoding efficiency. Meanwhile, the improvement of underlying system hardware, such as gradients[10], introduces another avenue for MRE development by enabling the simultaneous advancement of several areas described above.
Enhanced gradient performance in the form of improved peak amplitude and max slew rate are likely to facilitate improvements in scan time, SNR, and imaging artifacts via reductions in echo time, repetition time, and echo spacing, which improve phase acquisition. The potential benefits for elastography can be evaluated by implementing MRE on a high-performing gradient system such as the Microstructure Anatomy Gradient for Neuroimaging with Ultrafast Scanning (MAGNUS), with a slew rate of 500 m/T/s and a max amplitude of 200 mT/m[10]. MRE performance can be compared across the dynamic range of the gradient system to observe effects on imaging and to characterize shear wave behaviour, ultimately pointing towards areas for future MRE research with performance gradients.

Materials and

Methods: Agar-based MRE phantoms were built, characterized via mechanical testing, and scanned via MRE, followed by the collection of human scan data on the MAGNUS across gradient strength applications. Three custom MRE phantoms were set in cylindrical glass containers. The first was a homogeneous gel slurry made of 1.0 wt% Agar in water, stirred, heated to 84°C, and left to cool at room temperature. The second was a homogeneous hydrogel made of 0.5 wt% Agar in artificial cerebral spinal fluid (aCSF), autoclaved with a maximum temperature of 124.21°C and cooled at room temperature. The third phantom was an inhomogeneous hydrogel made of 0.5 wt% Agar in aCSF with the addition of two cylindrical stiff inclusions of 0.85 wt% Agar in aCSF, autoclaved and cooled in the same fashion as the second. Subsamples were taken from each phantom component for mechanical compression testing. All human data was collected according to an IRB-approved observational study.
All scans were performed in a GE MAGNUS using EPI-based MRE sequences and direct inversion-based processing pipelines (MR Touch and The Mayo Clinic pipeline). Parameters included actuation frequency, motion encoding gradient (MEG) frequency, actuation amplitude, and gradient amplitude. Optimal parameters were determined and sweeps across gradient amplitude were performed from a minimum of standard system operation (20 mT/m) up to the maximum (200 mT/m) to evaluate gradient strength effects derived from system capabilities. Octahedral shear strain SNR (OSS-SNR) and motion encoding efficiency (MENC) were recorded as performance indicators and the presence of phase wrap was noted.

Results, Conclusions, and Discussions: Results & Discussion
In every scanning case, application of a stronger gradient resulted in lower (more efficient) motion encoding efficiency. For the homogeneous slurry and hydrogel phantoms, MENC was improved via gradient amplitude by factors of 6.0 and 2.3, respectively, while retaining reasonably interpretable data. In human scanning, MENC was improved by a factor of up to 8.5.
Greater efficiency does not, by itself, mean better scan quality. For phantoms, enhanced efficiency alone often leads to artifacts from excessive phase wrap. This can be seen in the reduction of MENC and variation of OSS-SNR in the slurry phantom as gradient amplitude was increased while all other parameters are held constant. Despite a 6.8-fold improvement in MENC, severe phase wrap resulted in a reduction in OSS-SNR from a peak of 16.06 to 5.59. Parameter adjustment was easily applied to reduce artifacts and recover OSS-SNR. This effect was also observed in humans at extreme gradient strengths, although to a much lesser extent than in the phantoms due, most likely, to skull-brain motion transfer dynamics.
Phase wrap was most severe in the homogeneous hydrogel phantom, followed by the inhomogeneous hydrogel phantom and the slurry phantom. The structural composition of a material may affect its load-specific stiffness, or standing waves may contribute to greater phase accumulation. These observations warrant further exploration.
Conclusions
In conclusion, high-performance gradient systems enable enhanced motion encoding efficiency, thereby increasing flexibility in parameter configuration, scan time tradeoffs, signal acquisition, and more, which could result in higher SNR and resolution. However, this jump in efficiency entails more intentional protocol design and MRE method development in order to make use of the benefit without introducing phase wrap. Finally, differences observed in phantom structure types and their measured stiffness values under different loading conditions introduce new MRE characterization questions.

Acknowledgements (Optional): References
[1] R. Muthupillai, D. J. Lomas, P. J. Rossman, J. F. Greenleaf, A. Manduca, and R. L. Ehman, “Magnetic Resonance Elastography by Direct Visualization of Propagating Acoustic Strain Waves,” Science (80-. )., vol. 269, no. 5232, pp. 1854–1857, 1995, doi: 10.1126/SCIENCE.7569924.
[2] L. V. Hiscox et al., “Magnetic resonance elastography (MRE) of the human brain: technique, findings and clinical applications,” Phys. Med. Biol., vol. 61, no. 24, p. R401, Nov. 2016, doi: 10.1088/0031-9155/61/24/R401.
[3] M. C. Murphy, J. Huston, and R. L. Ehman, “MR elastography of the brain and its application in neurological diseases,” Neuroimage, vol. 187, pp. 176–183, Feb. 2019, doi: 10.1016/J.NEUROIMAGE.2017.10.008.
[4] R. L. Ehman, K. J. Glaser, and A. Manduca, “Review of MR elastography applications and recent developments,” J. Magn. Reson. Imaging, vol. 36, no. 4, pp. 757–774, Oct. 2012, doi: 10.1002/JMRI.23597.
[5] C. L. Johnson and E. H. Telzer, “Magnetic resonance elastography for examining developmental changes in the mechanical properties of the brain,” Dev. Cogn. Neurosci., vol. 33, pp. 176–181, Oct. 2018, doi: 10.1016/J.DCN.2017.08.010.
[6] H. Schwarb, C. L. Johnson, M. D. J. McGarry, and N. J. Cohen, “Medial temporal lobe viscoelasticity and relational memory performance,” Neuroimage, vol. 132, pp. 534–541, May 2016, doi: 10.1016/J.NEUROIMAGE.2016.02.059.
[7] C. L. Johnson et al., “3D multislab, multishot acquisition for fast, whole-brain MR elastography with high signal-to-noise efficiency,” Magn. Reson. Med., vol. 71, no. 2, pp. 477–485, Feb. 2014, doi: 10.1002/MRM.25065.
[8] C. Guenthner, J. H. Runge, R. Sinkus, and S. Kozerke, “Analysis and improvement of motion encoding in magnetic resonance elastography,” Nmr Biomed., vol. 31, no. 5, May 2018, doi: 10.1002/NBM.3908.
[9] D. R. Smith et al., “Anisotropic mechanical properties in the healthy human brain estimated with multi-excitation transversely isotropic MR elastography,” Brain multiphysics, vol. 3, Jan. 2022, doi: 10.1016/J.BRAIN.2022.100051.
[10] T. K. F. Foo et al., “Highly efficient head-only magnetic field insert gradient coil for achieving simultaneous high gradient amplitude and slew rate at 3.0T (MAGNUS) for brain microstructure imaging,” Magn. Reson. Med., vol. 83, no. 6, pp. 2356–2369, Jun. 2020, doi: 10.1002/MRM.28087.