Real-Time Multifrequency MR Elastography of the Human Brain Reveals Rapid Changes in Viscoelasticity in Response to the Valsalva Maneuver

Abstract

Modulation of cerebral blood flow and vascular compliance plays an important role in the regulation of intracranial pressure (ICP) and also influences the viscoelastic properties of brain tissue. Therefore, magnetic resonance elastography (MRE), the gold standard for measuring in vivo viscoelasticity of brain tissue, is potentially sensitive to cerebral autoregulation. In this study, we developed a multifrequency MMRE technique that provides serial maps of viscoelasticity at a frame rate of nearly 6 Hz without gating, i.e., in quasi-real time (rt-MMRE). This novel method was used to monitor rapid changes in the viscoelastic properties of the brains of 17 volunteers performing the Valsalva maneuver (VM). rt-MMRE continuously sampled externally induced vibrations comprising three frequencies of 30.03, 30.91, and 31.8 Hz were over 90 s using a steady-state, spiral-readout gradient-echo sequence. Data were processed by multifrequency dual elasto-visco (MDEV) inversion to generate maps of magnitude shear modulus | G∗| (stiffness) and loss angle φ at a frame rate of 5.4 Hz. As controls, the volunteers were examined to study the effects of breath-hold following deep inspiration and breath-hold following expiration. We observed that | G∗| increased while φ decreased due to VM and, less markedly, due to breath-hold in inspiration. Group mean VM values showed an early overshoot of | G∗| 2.4 ± 1.2 s after the onset of the maneuver with peak values of 6.7 ± 4.1% above baseline, followed by a continuous increase in stiffness during VM. A second overshoot of | G∗| occurred 5.5 ± 2.0 s after the end of VM with peak values of 7.4 ± 2.8% above baseline, followed by 25-s sustained recovery until the end of image acquisition. φ was constantly reduced by approximately 2% during the entire VM without noticeable peak values. This is the first report of viscoelasticity changes in brain tissue induced by physiological maneuvers known to alter ICP and detected by clinically applicable rt-MMRE. Our results show that apnea and VM slightly alter brain properties toward a more rigid-solid behavior. Overshooting stiffening reactions seconds after onset and end of VM reveal rapid autoregulatory processes of brain tissue viscoelasticity.

Microscopic multifrequency MR elastography for mapping viscoelasticity in zebrafish

“…Numbers of motion‐encoding gradients periods were 14, 17, and 19, corresponding to 1000, 1200, and 1400 Hz vibration frequencies, respectively. Although this frequency range is too limited for analysis of viscoelastic dispersion in zebrafish tissue, previous work has shown that even minimal variation of excitation frequency enhances the stability of the inverse problems solution for the reconstruction of MRE maps using our k‐MDEV (multifrequency dual elasto‐visco) algorithm 36,37 . For testing spatial resolution and SNR using the ultrasound gel phantom, matrix sizes of 100 × 100, 67 × 67, 50 × 50, and 40 × 40 were used with a 4 × 4 mm 2 FOV, resulting in in‐plane pixel sizes of 40 × 40 µm 2 , 60 × 60 µm 2 , 80 × 80 µm 2 , and 100 × 100 µm 2 , respectively.…”

Inversion‐recovery MR elastography of the human brain for improved stiffness quantification near fluid–solid boundaries

“…Similar empirical SWS thresholds were used in previous MRE studies, because larger fluid-filled regions and blood vessels that appear enlarged in SWS maps are efficiently eliminated. 23,25,26 The SNR was estimated as the ratio of the magnitude MRE signal between parenchyma and air. Additionally, we quantified the displacement SNR in wave images using the blind noise estimation method proposed by Donoho et al, 29 1.8 ± 0.3 m/s [IR-MRE], P = .21).…”

“…Brain‐surface ROIs were finally obtained by subtracting the eroded M from the original M. Mean parenchymal values were quantified over all voxels within M with SWS > 0.9 m/s. Similar empirical SWS thresholds were used in previous MRE studies, because larger fluid‐filled regions and blood vessels that appear enlarged in SWS maps are efficiently eliminated 23,25,26 …”

“…Vibration frequency was set to 31.25 Hz at 650 mbar (Figure 1C). The frequency range of 30 Hz has been exploited in other studies of brain MRE as a viable compromise between limited wave penetration at higher frequencies and limited wave number at lower frequencies 26 …”

Investigation of Soft Matter Nanomechanics by Atomic Force Microscopy and Optical Tweezers: A Comprehensive Review

“…The viscous response of soft matter samples is determined by the extension of G ″. Multifrequency MRE has been successfully devoted to address the elastic properties of agarose biopolymers [ 93 ], hydrogels [ 94 ], decellularized pancreatic tissues [ 95 ], brain tissues [ 96 ] and inflammatory bowel diseases [ 97 ], among others. Recently, multifrequency MRE setup was coupled with high-speed cameras to obtain ultrafast images of cellular elasticity [ 98 ].…”