4801

Analyzation and Optimization of Simultaneous Non-Contrast Angiography and intraPlaque Hemorrhage (SNAP) Magnetic Resonance Angiography in Intracranial Vascular Imaging
Yuhui Xiong1, Le He1, Yu Ma2, Xihai Zhao1, Chun Yuan1,3, and Hua Guo1

1Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, China, 2Tsinghua University Yuquan Hospital, Beijing, China, 3Vascular Imaging Laboratory, Department of Radiology, University of Washington, Seattle, WA, United States

Synopsis

In this study, we used theoretical simulation and in-vivo scan validation to explore the relationship between intracranial SNAP-MRA signal and blood flow velocity to explain why the performance of SNAP-MRA may vary greatly between healthy people and patients with neurodegenerative diseases. We also found the way to re-optimize the scan parameters (TI and flip angle) to enhance the SNAP-MRA signal and acquire high quality intracranial SNAP-MRA.

Purpose

Simultaneous Non-contrast Angiography and intraPlaque imaging (SNAP)1 is a carotid artery imaging technique which can provide a full 3D luminal magnetic resonance angiography (MRA) and an intrinsically registered 3D intraplaque hemorrhage (IPH) image in a single scan1. The SNAP-MRA can provide intracranial vascular images with effective background suppression offered by the phase-sensitive reconstruction technique3, and its potential in intracranial vascular stenosis detection has been shown2. However, we found that sometimes the performance of intracranial SNAP-MRA may vary greatly between healthy people and patients with neurodegenerative diseases (Fig. 1). The blood flow velocity may be the potential cause and it was mentioned without theoretical analysis in previous studies1, 4. Moreover, the scan parameters (TI and flip angle) in the commonly used SNAP sequence were optimized for maximized carotid IPH-vessel wall and vessel wall-lumen contrasts1, while in intracranial SNAP-MRA we are more concerned about the intensity of blood signal. This study aims to explore the relationship between SNAP-MRA signal and blood flow velocity and re-optimize the intracranial SNAP-MRA sequence parameters to enhance the SNAP-MRA signal using theoretical simulation and in-vivo scan validation.

Methods

Simulation In the simulation conducted in Matlab (R2016a, Mathworks, Natick, MA), we fixed some sequence parameters (shown in Fig. 2a): TR=10 ms, Tgap=15 ms, Tex=5 ms, β=5°, IRTR=4*TI, IR slab thickness TIR=62 cm, imaging slab thickness Tsl=6 cm. The T1 and T2 for different tissues5, 6 are listed in Fig. 2c. Firstly we simulated the ideal SNAP signal evolution for the brain (Fig. 2b) with conventional scan parameters: TI=500 ms, α=11°. The following simulations were based on assumptions that blood flow is advection with a constant velocity, and the MR signal right after TI was regarded as the SNAP signal. We selected several different TI values ranging from 200 ms to 600 ms, and calculated the corresponding SNAP-MRA signal intensity with blood flow velocity ranging from 5 cm/s to 150 cm/s. For sequence optimization, TI values between 200-1000 ms and α between 5-25° were attempted.

In-vivo validation The scans were performed on a Philips 3.0T Achieva TX MR scanner (Philips Healthcare, Best, The Netherlands) using an 32-channel SENSE head coil. Data were acquired from one healthy volunteer. Multiple SNAP sequences with different TI and α were scanned. A PPU-triggered 2D QFLOW sequence based on phase-contrast imaging was scanned to measure the actual blood flow velocity. The detailed scan parameters were listed in Fig. 5a. SNAP images were zero-padded in all directions to achieve 0.4 mm isotropic resolution. The SNAP-MRA images after phase-sensitive reconstruction and maximum intensity projection (MIP) were compared with each other to validate the conclusions of the simulation.

Results and Discussion

The simulation results of SNAP-MRA signal at different blood flow velocities are shown in Fig. 3. As the fast blood flow outside the IR slab may flow into the image slab without being inversed and the slow blood flow may be inversed for more than once before imaging, only the blood flow with velocities within an “interval” can generate a steady signal. The range of the “velocity interval” can be written as:

$$V\geq\frac{T_{IR}+T_{sl}}{2\cdot(IRTR+TI)}=\frac{62cm+6cm}{10\cdot{TI}}=\frac{6.8cm}{TI}\qquad{and}{\qquad}V\leq\frac{T_{IR}-T_{sl}}{2\cdot{TI}}=\frac{62cm-6cm}{2\cdot{TI}}=\frac{28cm}{TI}$$

which is consistent with previous work1, 4. Blood flow velocity outside the “velocity interval” will result in images like Fig. 1b. The simulation result of SNAP-MRA signal with different scan parameters (Fig. 4) shows that the SNAP-MRA signal can be enhanced by selecting a smaller TI and flip angle within a reasonable range.

The in-vivo SNAP-MRA images are shown in Fig. 5b. The 2D QFLOW sequence showed that the blood flow velocity through MCA was 17.3 cm/s. When TI=220 ms, this velocity is outside the “velocity interval” and its corresponding image contains few vessel branches like Fig. 1b. The image intensity of the third image is obviously higher than that of the first image and shows better image details (yellow arrowheads). Image intensity from the same ROI (yellow dotted box) of three images were added up respectively and compared, resulting in 1: 2.29: 1.42, which is highly consistent with the simulation results (marked in Fig. 4, 1: 2.34: 1.38).

Many previous works7-10 have found cerebral hemodynamic changes in neurodegenerative diseases. By scanning with appropriate parameters, intracranial SNAP-MRA can screen people with abnormal cerebral blood flow velocity, making it be potential in diagnosing of neurodegenerative diseases like Alzheimer’s disease and Parkinson’s disease.

Conclusion

This study shows the effects of blood flow velocity, TI and flip angle on intracranial SNAP-MRA. To acquire high quality intracranial SNAP-MRA images, TI should be determined according to the blood flow velocity, and flip angle can be determined subsequently for high SNAP-MRA signal intensity.

Acknowledgements

No acknowledgement found.

References

[1] Wang J, Börnert P, Zhao H, et al. Simultaneous noncontrast angiography and intraplaque hemorrhage (SNAP) imaging for carotid atherosclerotic disease evaluation[J]. Magnetic resonance in medicine, 2013, 69(2): 337-345.

[2] Wang J, Guan M, Yamada K, et al. In Vivo Validation of Simultaneous Non-Contrast Angiography and intraPlaque Hemorrhage (SNAP) Magnetic Resonance Angiography: An Intracranial Artery Study[J]. PloS one, 2016, 11(2): e0149130.

[3] Kellman P, Arai A E, McVeigh E R, et al. Phase-sensitive inversion recovery for detecting myocardial infarction using gadolinium-delayed hyperenhancement[J]. Magnetic resonance in medicine, 2002, 47(2): 372-383.

[4] Wang J, Ferguson M S, Balu N, et al. Improved carotid intraplaque hemorrhage imaging using a slab-selective phase-sensitive inversion-recovery (SPI) sequence[J]. Magnetic resonance in medicine, 2010, 64(5): 1332-1340.

[5] Noeske R, Seifert F, Rhein K H, et al. Human cardiac imaging at 3 T using phased array coils[J]. Magnetic resonance in medicine, 2000, 44(6): 978-982.

[6] Wansapura J P, Holland S K, Dunn R S, et al. NMR relaxation times in the human brain at 3.0 tesla[J]. Journal of magnetic resonance imaging, 1999, 9(4): 531-538.

[7] Nobili F, De Carli F, Frisoni G B, et al. SPECT predictors of cognitive decline and Alzheimer's disease in mild cognitive impairment[J]. Journal of Alzheimer's Disease, 2009, 17(4): 761-772.

[8] Binnewijzend M A A, Kuijer J P A, Benedictus M R, et al. Cerebral blood flow measured with 3D pseudocontinuous arterial spin-labeling MR imaging in Alzheimer disease and mild cognitive impairment: a marker for disease severity[J]. Radiology, 2013, 267(1): 221-230.

[9] Klohs J, Rudin M, Shimshek D R, et al. Imaging of cerebrovascular pathology in animal models of Alzheimer's disease[J]. Frontiers in aging neuroscience, 2014, 6.

[10] Kara F, van Dongen E S, Schliebs R, et al. Monitoring blood flow alterations in the Tg2576 mouse model of Alzheimer's disease by in vivo magnetic resonance angiography at 17.6 T[J]. Neuroimage, 2012, 60(2): 958-966.

Figures

Fig. 1. 3D SNAP-MRA images after phase-sensitive reconstruction and MIP. Image a was acquired from a healthy volunteer (24y, male) while image b was acquired from a patient (77y, male) with Alzheimer’s disease. Two images were acquired under the same scan parameters and were displayed using the same window levels. It is very obvious that in image b most of the branch vessels are not shown or with low image intensity.

Fig. 2. a, Schematic illustration of the SNAP sequence. It’s an inversion recovery magnetization prepared gradient echo sequence, followed by a series of small flip angle acquisition serving as the reference1. b, Simulation of the ideal SNAP signal evolution for the brain with conventional scan parameters: TI=500ms, α=11°. c, The T1 and T2 relaxation times for different tissues5, 6.

Fig. 3. Simulation of SNAP-MRA signal at different blood flow velocities. The flip angle was fixed as 11° in this simulation. The horizontal axes represent blood flow velocity in cm/s, while the vertical axes represent the blood signal intensity. The SNAP-MRA signal within the “velocity interval” is stable while the signal outside the interval reduced to zero rapidly. The small sidelobes in the first six panels come from the low velocity blood flow that has experienced multiple (≥3) IR pulses before imaging. In each panel, the starting and ending points of the “velocity interval” are marked.

Fig. 4. Simulation of SNAP-MRA signal with different scan parameters. The lower left and upper right portion of this image is removed to ensure the signals of other tissues (gray matter, white matter and CSF) keep positive while the blood signal keeps negative, leading to effective background suppression and integrated SNAP-MRA image. It is worth noting that the absolute value of pixel intensity in this image is proportional to the corresponding SNAP-MRA signal intensity. The scan parameters used in the in-vivo validation and corresponding simulated signal intensity are marked.

Fig. 5. Validation of “velocity interval” and sequence optimization using in-vivo scans. a, The detailed scan parameters. The scan time of the QFLOW sequence was about 4 minutes when the heart rate of volunteer was about 52/min. b, 3D SNAP-MRA images with different scan parameters. Images were acquired from a healthy volunteer (25y, female) and were displayed using the same window levels. The scan parameters of each image and corresponding “velocity interval” were displayed above and below the image, respectively. Yellow dotted box in the second image is the ROI used in quantitative intensity comparison of three images.

Proc. Intl. Soc. Mag. Reson. Med. 26 (2018)
4801