Three-dimensional (3D) time-of-flight MR angiography (TOF MRA) has been widely used. In TOF MRA, complex or turbulent flow, such as carotid artery bifurcation, causes phase dephasing artifacts that lead to an intraluminal signal decrease or a loss of signal. Silent MRA is an imaging technique using arterial spin labeling that make it possible to zero echo time (TE) by applying 3D radial acquisition and ultra-fast switching of the magnetic gradient [1]. A previous report has shown Silent MRA was able to visualize flow in an intracranial stent more effectively than TOF MRA [2]. Silent MRA is expected to decrease the effect of intra-voxel dephasing for turbulent flow and allow for accurate assessment of flow in blood vessels. This study quantified the accuracy of Silent MRA for visualizing turbulent flow in flow-phantom and in vivo studies.

The MRA studies were performed on a 3.0T Discovery 750w MRI system with a 12 channel phased array coil (GE Healthcare, Waukesha, Wisconsin).

Phantom study

Flow phantoms were 5.0 mm inner diameter polyvinyl tubes with and without stenosis that were submerged in polyvinyl alcohol gel (T1 = 1447 ms). The flow phantoms were connected to a pump with polyvinyl tubing. A blood analog (30% glycerin and 70% distilled water, T1 = 1400 ms) mimicking the viscosity and T1 relaxation time of blood was circulated with steady flow at a velocity of 20 and 40 cm/s, respectively, through the phantoms (Fig. 1). MRA images were obtained using TOF and Silent MRA with and without stenosis at both flow velocities. The imaging parameters were: field of view (FOV) = 200 mm, repetition time (TR)/TE = 20/2.7 ms, Flip angle (FA)[A2] = 20°, matrix size = 416 x 192, slice thickness = 1.2 mm, bandwidth (BW) ± 42 kHz. Silent MRA uses a continuous arterial spin labeling technique as a preparation pulse, and a 3D radial acquisition [1]. Silent MRA in the sagittal plane was acquired with control images and labeling images, and they were subtracted for the MRA image. The imaging parameters were: FOV = 180 mm, TR/TE = 1114/0.016 ms, FA = 5°, spokes per segment = 512, resolution = 1.2 mm, BW ± 20 kHz. Signal intensity was measured in the tube for a normal model and in the distal region that passed through the stenosis for the stenosis model. Then, the image contrast between the signal intensity of the tube and peripheral region were measured at flow velocities of 20 and 40 cm/s, with and without stenosis. The coefficient of variation (CV) in the tube was measured using the same conditions with the previous settings.

In vivo experiments

Seven healthy subjects (mean age 38.7 years) were enrolled under an Internal Review Board-approved protocol. TOF and Silent MRA images were obtained with the same imaging parameters as used in the phantom study. Next, signal intensities were measured for each MRA image at the level of the S1, S3, S4, and S5 segment of the internal carotid artery and at the S4 level of the segment basilar artery. The contrast between each vessel and the peripheral brain parenchyma were measured in TOF and Silent MRA. Moreover, signal intensity profiles were measured along directions perpendicular to vessels at the level of the S3 and S5 segment. Finally, to assess signal changes in the blood vessel, an accuracy that was defined as the ratio normalized maximum signal for minimum signal was measured.

In the phantom study, although the contrast of TOF MRA decreased with high flow velocity or distal to the stenosis region, that of Silent MRA was not changed by these conditions (Fig. 2a). Moreover, the mean CV of TOF MRA was higher than that of Silent MRA in the stenosis condition (Fig. 2b). These results are consistent with a previous report [3]. In the in vivo study, the mean contrast of Silent MRA was statistically significantly higher than that of TOF MRA in all segments (Fig. 3). Representative MRA images and the signal intensity profile of TOF and Silent MRA are shown in Figure 4. Moreover, the accuracy of Silent MRA was statistically significantly higher than that of TOF MRA in the S3 and S5 segments (P < 0.05) (Fig. 5).Silent MRA improved contrast, CV, and accuracy for intracranial blood vessels with turbulent flow. Signal intensities obtained by Silent MRA were independent of flow conditions. Although it was limited in spatial resolution and required additional imaging time, it may have the potential to improve uniformity of signal intensity of intracranial vessels.