Synopsis

We confirmed that 2D beam excitation presaturation-pulse (hereafter BeamSat pulse) can saturate the vessels selectively in brain. Although BeamSat can visualize hemodynamics, the scan time is long (about 5 minutes) because the based sequence is 3D TOF. The single thick slice 2D TOF was used for based sequence to reduce the scan time and the k_s offset was added to saturate background brain signal. We showed that this method can accelerate the scan time to 0.6-0.8 seconds without the notable loss of vessel visualization.

Introduction

Materials and Methods

Imaging sequence

BeamSat pulse was set prior to 2D TOF sequence in each TR to suppress the inflow signal from the internal carotid artery (ICA) (Figure 1). To suppress ICA selectively, BeamSat pulse was set in the petrous portion of ICA (Figure 2a-b). The number of slice was one, and slice selective region was from MCA-M1 to M4 (Figure 2a). The k_s offset gradient was added in slice direction to select the targeted k-space (Figure 1). Other typical scan parameters: TR/TE = 30.3 ms/6.9ms, imaging matrix 512 / 320, Thickness = 50 [mm], rectFOV = 85%, acceleration factor = 2.0, scan time = 4.1 [sec]. The control 2D TOF MRA without BeamSat pulse was obtained using same scan parameter and k_s offset gradient. The subtracted image of control and 2D TOF with BeamSat was obtained in each k_s offset (Figure 3). A 1.5T MRI scanner and 8 channel Head coil were used for all imaging. A healthy volunteer was imaged. We explained the purpose and significance of this study to healthy volunteers and obtained written consent.

Optimization of k_s offset and k_p sampling reduction

We optimized the k_s offset and acquiring phase data region (Wk_p). First, the k_s offset was changed from 0 [m^-1] to 400 [m^-1]. The visibility of blood vessels and residual brain signal was evaluated by subtracted images. Second, optimized k_s was used, and Wk_p was reduced from 10% to 25% with a focus on k_p = 0 [m^-1]. The visibility of blood vessels was evaluated by subtracted images.

Results and Discussion

Optimization of k_s offset

In ks = 0 [m^-1] (Figure 4a), the residual brain signal was visualized, and blood vessels were not visualized clearly, so simple subtraction was not enough to saturate the background brain signal. In k_s = 60 [m^-1], blood vessels were visualized clearly without residual brain signal (Figure 4b). When the general size of brain and vessels (diameter) were 50-100 [mm] and 1-3 [mm], main bandwidth of brain and vessels were 20-40 [m^-1] and 667-2000 [m^-1] in the k-space (Figure 5). It is supposed that optimal k_s was minimum value more than 40 [m^-1]. The actual optimized k_s offset value (60 [m^-1]) was comparable with that. The scan time was accelerated to 4.1 [sec]. Although it was shorter than conventional selective TOF MRA (5 minutes), 4.1 [sec] was long to visualize the time-dependent change like a DSA.

Optimization of k_p sampling reduction

The blood vessels were not visualized clearly in the image of Wk_p = 10%, and these were visualized in the image of Wk_p = 15-20% (Figure 4c-f). Scan time of Wk_p = 15-20% was accelerated to 0.6-0.8 [sec], and it was not so long for time frame. In order to accelerate the scan time more, each frame should be obtained individually. The delay can be controlled by changing the time applying BeamSat pulse continuously before the data acquisition. When each frame was visualized individually, the time to remove the saturated blood was needed after the data acquisition. The removal time is almost 2 [sec]. When maximum delay time is 2 sec, the time of each frame acquisition is 4.6-4.8 [sec] that is sum of delay time (2 [sec]) and acquisition time (0.6-0.8 [sec]) and removal time (2 [sec]). It may be useful for multi-phase scanning.

Conclusion

Acknowledgements

References

1. Nishihara T et al. Proc In 20th ISMRM, Selective TOF MRA using Beam Saturation pulse. 2012;2497.

2. Ito K, Sasaki M et al. Noninvasive Evaluation of Collateral Blood Flow through Circle of Willis in Cervical Carotid Stenosis Using Selective Magnetic Resonance Angiography. J Stroke Cerebrovasc Dis. 2014;23(5):1019-23.

3. Sodickson A, Mortele KJ, Barish MA, et al. Threedimensional fast-recovery fast spin -echo MRCP: comparison with two-dimensional single -shot fast spin-echo techniques. Radiology 2006;238:549-559.