record three orthogonal displacement components (u1, u2, u3) relative to the oblique imaging plane. The absolute peak motion-encoding gradient amplitude was 25 mT/m, nominally, for all test frequencies and spin-phase was accrued over a single gradient cycle. All data were acquired with a temporal resolution of four points per actuation cycle. Total scan time per subject was approximately 21 minutes. In this time, all motion components were acquired and for each actuation frequency. Figure 4: (Left) An active acoustic driver produced an oscillatory pressure wave which was transmitted to each passive actuator pad via flexible tubing. Passive actuator pads were positioned on the subject’s skull near the left and right pterion bones and affixed via elastic bandage (not shown). (Right) Acoustic pressure was measured via transducer embedded in the tubing. A typical waveform is shown. Motion was induced in the brain using an acoustic actuation system (Resoundant™, Resoundant Inc.) modified slightly so that a single active driver could power two passive drivers with equal amplitude and phase. Each passive actuator was positioned on the subject’s skull near the left and right pterion bones and affixed with an elastic bandage (not shown), cf. Figure 4. The actuation system was configured to transmit a 4, 6, and 8 cycle pressure-wave train synchronized with the MRE sequence at 45, 60 and 80 Hz, respectively. Imposed acoustic pressure loads were measured with the PCB Piezotronics model 103B01 dynamic pressure sensor. Data processing Motion encoded MR data were obtained using switched-polarity encoding acquisition scheme to remove systematic phase errors and enhance displacement contrast. This scheme increases the displacement contrast by a factor of two since two motion-encoded images, each with opposite polarity, are required. Phase-contrast images were obtained by complex division of positive and negative polarity phase images, and converted into units of length. Sensitivity factors of 5.63 (45 Hz), 7.66 (60 Hz), and 10.6 μm/rad (80 Hz) data were calculated by numerical integration of Equation 1 – taking into consideration actual gradient performance specifications. Data were filtered with a 3 x 3 kernel median filter and circular, 4th-order Butterworth bandpass filter (high cut: 2.14 cm, low cut: 40.0 cm). Displacement gradients and shear angles were observed to be small for all experiments performed; therefore, the 2-D infinitesimal strain tensor was calculated, ߝ ൌଵ ଶ ൫ ݑ ǡ ݑ ǡ൯ ሺ݅ ǡ݆ ൌ ͳǡ ʹሻ. (3) Shear-strains were evaluated in terms of the root-mean-squared (RMS) scalar quantity and normalized by the RMS applied pressure load. Pressure-normalized RMS shear-strains were evaluated in two fashions for each subject: i.) globally – across all brain matter, and ii.) regionally – in three concententric ring-like zones, in hopes of illuminating the dispative properties of brain tissue. The imaging procedure was repeated three times with different motion-encoding, magnetic field gradient orientations to 61
RkJQdWJsaXNoZXIy MTMzNzEzMQ==