Home Computer Science Computational Diffusion MRI: MICCAI Workshop, Athens, Greece, October 2016
Trapezoidal OGSE diffusion sequences, as shown in Fig. 2, are implemented on a Philips Achieva 3.0T TX MRI system (University College London Hospital, London, UK). We choose trapezoidal OGSE waveforms with a fixed maximum gradient strength as it has been shown previously that these are the most sensitive to microcapillary diameters [19, 23]. The main user controlled parameters are echo time (TE), pulse duration (1), diffusion time (A) and number of half period oscillations, referred to as ‘lobes’ (N). Gradient strength, G, and slew rate for the trapezoid waveforms are fixed at 62mT/m and 68.9mT/m/ms, respectively, to adhere to manufacturer set threshold for peripheral nervous stimulation (PNS). The b-value for the OGSE sequences with trapezoidal gradient are calculated as in .
The plates are scanned during the same session using Philips SENSE Flex Surface coils. A room temperature of 20 °C is maintained throughout the experiment. The diffusion protocol consists of nine HARDI shells with b-values 12020,000 s/mm2, each with 32 gradient directions and one b=0s/mm2. The shells have a fixed pulse duration (1 = 39 ms, A = 63 ms) but the number of lobes varies from N = 1 to N = 9 (i.e. frequencies between 12.8-115 Hz), and consequently the
Fig. 2 Schematic representation of the OGSE diffusion imaging protocols (left) and corresponding plate example images (right). The protocol included (a) the single shot echo planar imaging (SS-EPI) sequence containing the excitation, refocusing pulse and readout timings; and the OGSE sequences with (b) N = 1, (c) N = 2 and (d) N = 9. The parameters depicted here are: echo time (TE), pulse duration (1), diffusion time (A), gradient strength (G) and number of lobes (N). The example plate images show the 5 p,m pair (immersed vertically in water) scanned perpendicular to the plane of the plate. (e) is the non-diffusion weighted image. (f), (g) and (h) display diffusion weighted images for OGSE sequence shown in (b), (c) and (d) respectively. The diffusion weighted images are in the parallel and two nearly perpendicular directions relative to the long axis of the microcapillaries. These are only three example directions of the 32 gradient directions that were used in this study. High signal attenuation is seen in the parallel gradient direction indicating free diffusion of water along the long axis of the microcapillaries. Signal appears bright in the perpendicular directions which comes from the restricted diffusion of water across the long axis of the microcapillaries
b-values varied (see Fig. 2). An additional, standard SDE diffusion sequence (N = 1, 1 =10 ms, A = 92 ms) with a b-value of 1860 s/mm2 is also included for comparison.
All diffusion protocols use single-shot-echo-planar imaging (SS-EPI). Each acquired image has one slice of thickness 10 mm, which is orthogonal to the plane of the plate (see Fig. 2). The imaging matrix is 76 x 19 with a resolution of 0.4 x 1.6 mm, which is used to ensure at least one row of the voxels does not contain partial volume effects. In order to obtain sufficient diffusion weighting for all N, we extend the diffusion gradient duration by using a long echo time (in terms of clinical scanning) TE= 120 ms for all shells. Other sequence parameters are: Half Fourier = 0.8, TR = 3 s, repetitions = 1 and acquisition time per protocol =1.75 min.
|< Prev||CONTENTS||Next >|