Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Dec 1.
Published in final edited form as: Magn Reson Med. 2020 Jun 10;84(6):3316–3324. doi: 10.1002/mrm.28339

High Spatial Resolution Whole-Neck MR Angiography using Thin-Slab Stack-of-Stars Quiescent Interval Slice-Selective (QISS) Acquisition

Ioannis Koktzoglou 1,2, Rong Huang 1, Archie L Ong 2,3, Pascale J Aouad 1,4, Matthew T Walker 1,2, Robert R Edelman 1,4
PMCID: PMC7722209  NIHMSID: NIHMS1627958  PMID: 32521094

Abstract

Purpose:

To report a 3D multi-echo thin-slab stack-of-stars (tsSOS) quiescent-interval slice-selective (QISS) strategy for high-resolution magnetic resonance angiography (MRA) of the entire neck in under seven minutes.

Methods:

The neck arteries of 8 subjects were imaged at 3 Tesla. Multi-echo 3D tsSOS QISS using a FLASH readout was compared with 3D tsSOS FLASH, 2D QISS, 2D TOF, and 3D TOF. A root-mean-square (RMS) combination of echo time images was tested. Evaluation metrics included arterial signal-to-noise ratio (SNR), arterial-to-muscle contrast-to-noise ratio (CNR), and image quality.

Results:

3D multi-echo tsSOS QISS using a RMS combination of echo time images increased SNR and CNR by 60% and 63% with respect to the reconstruction obtained with the shortest echo time. 3D tsSOS QISS showed superior CNR with respect to 3D tsSOS FLASH imaging, and more than 3-fold higher SNR and CNR with respect to 2D radial QISS when normalized for voxel size. 3D tsSOS QISS provided good to excellent image quality that exceeded the image quality of 2D QISS, 2D TOF, and 3D TOF (P<0.05).

Conclusion:

Whole-neck high-resolution nonenhanced MRA is feasible using 3D tsSOS QISS, and produced image quality that exceeded those of competing nonenhanced MRA protocols at 3 Tesla.

Keywords: 3D QISS, MRA, neck, carotid, vertebral, stack of stars

INTRODUCTION

Contrast-enhanced magnetic resonance angiography (CEMRA) is a routine diagnostic tool for detecting arterial disease in the neck (1,2). However, gadolinium-based contrast administration is contraindicated in patients with severe renal insufficiency (3), is associated with gadolinium deposition within the brain and body (4–6), and can interfere with contrast-enhanced perfusion MRI of the brain when performed as part of an acute stroke MRI protocol (7). 2D and 3D time of flight (TOF) techniques provide a risk-free nonenhanced alternative to CEMRA (8–11). 2D TOF can be used to efficiently survey vessels through the entire neck whereas, given longer scan times, 3D TOF is usually restricted to a small region through the carotid bifurcation. Key drawbacks of 2D and 3D TOF are suboptimally thick slices and limited anatomical coverage, respectively. Both approaches also apply large flip angle radiofrequency (RF) excitations to maximize arterial-to-background contrast, which can result in flow saturation effects.

Recent work has shown that 2D quiescent interval slice selective (QISS) MRA using radial k-space sampling provides better image quality than 2D TOF and good agreement with respect to contrast-enhanced MRA when evaluating the neck (12,13). Compared to TOF, 2D QISS reduces sensitivity to saturation artifacts (as smaller flip angles are used and the degree of background suppression is more decoupled from the flip angle), motion and pulsation artifacts, and more readily supports short scan times without loss of spatial resolution through the use of radial undersampling (14). Nonetheless, the slice resolution of 2D QISS is limited to approximately 2.0 mm due to gradient and RF pulse constraints, while the in-plane voxel dimensions are kept to approximately 1.0 mm or larger to maintain SNR (15).

Edelman et al. recently showed that a 3D thin-slab stack-of-stars (tsSOS) implementation of QISS provides improved resolution in the slice direction than 2D QISS for evaluating the lower extremity and renal arteries (16). In this pilot feasibility study, we hypothesized that a 3D tsSOS implementation of QISS using a multi-echo fast low-angle shot (FLASH) readout at 3T could provide better image quality for imaging of the neck vessels than is possible using existing nonenhanced MRA techniques, including the standard approaches of 2D and 3D TOF.

METHODS

This study was conducted under an institutional review board-approved research protocol. Imaging was performed on a 3 Tesla MRI scanner (MAGNETOM SkyraFit, Siemens Healthineers, Erlangen) equipped with a 20-channel head and neck coil. 8 human subjects were involved in this study, consisting of 7 healthy subjects and 1 patient with a history of bilateral carotid artery stenosis (5 M, 3 F, mean age 40.0±13.1 years). Written informed consent was obtained from all subjects prior to participation in the study.

3D Thin-Slab Stack-of-Stars QISS MRA

QISS refers to a class of non-contrast enhanced MRA techniques which apply dedicated RF pulses for in-plane and venous suppression, a quiescent interval (QI) for inflow of unsaturated arterial spins into the imaging slice or slab, followed by an imaging readout (17,18). Unlike prior reports of QISS for the neurovascular system, which utilized a 2D radial FLASH readout (12–14,19,20), the prototype 3D QISS technique applied here used a thin-slab stack-of-stars (tsSOS) multi-echo FLASH acquisition as shown in Figure 1a. Stack-of-stars sampling was done using an “inner slice loop” schedule (acquiring every slice-encoding step before acquiring the next radial view angle) which is beneficial for reducing transient and cyclical motion artifacts (21). The use of a FLASH acquisition also differs from the seminal 1.5 Tesla tsSOS study (16) which used balanced steady-state free precession and fast interrupted steady-state readouts (22,23).

Figure 1.

Figure 1.

Open in a new tab

Timing diagram and image reconstruction methodology. (a) Timing diagram of the 3D tsSOS QISS MRA pulse sequence. Following the application of in-plane and superior tracking inversion RF pulses for background and venous signal suppression, a 90 view 3-echo FLASH readout is applied. Gr and Gz denote the radial readout and slice-selection gradients, respectively. (b) Approach for root-mean-square (RMS) combination of TE images; images shown were acquired with 3D tsSOS QISS.

Whole-neck coverage was achieved by acquiring 19 thin overlapping slabs (slab thickness of 19.5 mm, 4.5 mm slab overlap) providing an axial coverage of 289 mm in 6 minutes 39 s. QISS suppression of background signal was done using an inversion pulse (thickness 19.5 mm) applied over the imaging slab. Imaging parameters for 3D tsSOS QISS are provided in Table 1. Slice-encoding steps were acquired in an ascending (i.e. non-centric) order, with the radial view angle incremented by 15.8245 degrees. Using 20% oversampling in the slice-encoding direction, a total of 14 (of 18) slice-encoding steps (6/8th partial Fourier) were acquired in each slab, from which (after 2-fold zero filling interpolation) the central 30 of 36 reconstructed slices were retained in each slab. A bipolar radial readout gradient was used to sample three TEs of 1.6 ms, 3.7 ms, and 5.7 ms. Ninety RF pulses (i.e. FLASH TRs) were acquired in each QISS TR. The acquisition window duration was 891 ms and the QI was 583 ms. Veins were suppressed using a 10-cm-thick tracking inversion RF pulse applied above each 3D slab. Fat saturation was not applied to avoid unintentional saturation of arterial spins due to magnetic field inhomogeneity (24). Relative to arterial signal, fat signal was suppressed through the combination of a low flip angle readout which lessened T1 weighting, the application of an in-plane inversion RF pulse, and data sampling at TEs where water and fat were predominantly out-of-phase.

Table 1.

Imaging Parameters

3D tsSOS QISSa 3D TOF 2D TOF 2D QISS
Orientation axial axial axial oblique axialb
Acquisition Type 3D stack-of-stars 3D Cartesian 2D Cartesian 2D radial
TR (ms) 9.9 21.0 23.0 15.0
TE (ms) 1.6, 3.7, 5.7 3.1 4.2 2.1
Asymmetric Echo yes yes yes
QISS TR/QI (ms) 1500/583 -- -- 1100/59
QISS contrast inversion-recovery -- -- inversion-recovery
Flip angle (degrees) 12 15 60 30
TONE excitation yes yes no no
Field of view (mm) 300 × 300 218 × 218 210 × 210 416 × 416
Matrix 352 × 352 256 × 256 256 × 256 384 × 384
Slabs Acquired 19 19 -- --
Slices Acquired 444 444 128 128
In-plane resolution (mm)c 0.85 × 0.85
(0.42 × 0.42)		 0.85 × 0.85
(0.42 × 0.42)		 0.82 × 0.82
(0.41 × 0.41)		 1.08 × 1.08
(0.54 × 0.54)
Slice thickness (mm)c 1.3 (0.65) 1.3 (0.65) 3.0 2.0
Acquired voxel size (mm3) 0.94 0.94 2.02 2.35
Slab thickness (mm) 19.5 19.5 -- --
Slice/slab overlap (mm) --/4.5 --/4.5 1.0/-- 0.6/--
Axial coverage (mm) 289 289 256 254
Phase Partial Fourier -- 6/8 off --
Slice Partial Fourier 6/8 6/8 -- --
Slice Oversampling (%) 20 20 -- --
Parallel Imaging Factor -- 5d 2 --
Flow Compensation no yes yes no
Bandwidth (Hz/pixel) 592 592 219 942
Scan time (s) 399 398 429 414
Open in a new tab
a

Parameters for 3D tsSOS FLASH are identical to those for 3D tsSOS QISS except that no inversion pulse was applied.

b

45° slice tilt between axial and coronal planes.

c

Values provided as: acquired (reconstructed).

d

Value chosen to the match scan time, anatomical coverage, and spatial resolution of 3D tsSOS QISS. QI = quiescent interval.

Comparisons with Other Nonenhanced MRA Protocols

Comparisons of whole-neck 3D tsSOS QISS were made with respect to 2D TOF, 3D TOF, 2D QISS protocols providing similar scan times of ≈7 minutes for imaging of the entire neck (respective axial coverages of 256 mm for 2D TOF, 289 mm for 3D TOF, and 254 mm for 2D QISS). Imaging parameters for these protocols are provided in Table 1. 3D TOF was acquired with the same spatial resolution, scan time, receiver bandwidth, slab thickness, number of imaging slabs, and spatial coverage as 3D tsSOS QISS, using a TE of 3.1 ms (which is the default value for our clinical protocols). A parallel imaging factor of 5 was used during 3D TOF to match the scan time, anatomical coverage, and spatial resolution of 3D tsSOS QISS. 2D radial QISS was performed as previously described (12,13), with each slice acquired in 3 shots. To determine the impact on image contrast of applying a preparatory inversion RF pulse, 3D tsSOS QISS was compared to 3D tsSOS FLASH using otherwise identical acquisition parameters. The acquisition order of the above protocols (3D tsSOS QISS, 3D tsSOS FLASH, 2D TOF, 3D TOF, 2D QISS) was randomized; no cardiac gating was used for any protocol.

In the patient study, an additional 3D TOF sequence (axial coverage of 126 mm) was acquired using our default clinical protocol. Imaging parameters for this protocol were: eight 19.2 mm thick slabs with 4.0 mm slab overlap, TR/TE/flip angle=20.0 ms/3.1 ms/15 degrees, FOV of 200 mm × 151 mm, 0.52 mm × 0.74 mm in-plane resolution, 1.6 mm slices interpolated to 0.8 mm, parallel imaging factor of 2, receiver bandwidth of 250 Hz/pixel, scan time of 3 min 49 s.

Image Processing of 3D tsSOS Data

To minimize venetian blind artifact at the slab boundaries, overlapping slice data from the 3D tsSOS slabs was combined through linearly-weighted averaging based on proximity to slab center. Two reconstructions from the 3D tsSOS volumes were evaluated, with one corresponding to the image set acquired with the shortest TE of 1.6 ms, and a second reconstruction obtained by root-mean-square (RMS) combination of all three acquired TEs (Figure 1b). RMS combination of multiple TEs was applied as it has previously been shown to improve signal-to-noise ratio (SNR) and accentuate T2* contrast in gradient-echo-based spine and musculoskeletal MRI (25–27), and could provide similar SNR and contrast benefits for 3D tsSOS MRA.

Quantitative Image Analysis

Arterial SNR and arterial-to-muscle contrast-to-noise ratio (CNR) were measured in six healthy subjects. SNR and CNR were measured using the dual acquisition subtraction method (28), through the acquisition of additional single-slab (for 3D protocols) or single-slice (for 2D protocols) acquisitions collected through the distal common carotid artery. Arterial SNR and arterial-to-muscle CNR were computed as Sa/σm and (Sa-Sm)/σm, where Sa and Sm denote mean signals in the common carotid artery and nearby sternocleidomastoid muscle, and σm is the noise level in sternocleidomastoid muscle. To account for protocol differences in spatial resolution, SNR and CNR values were normalized to a common voxel size of 1 mm3 prior to comparison, by dividing non-normalized values by the product x*y*z, where x, y, and z are the acquired voxel dimensions in millimeters.

Qualitative Image Analysis

Image volumes were loaded into a workstation (Leonardo, Siemens Healthineers, Erlangen) where non-arterial background was cropped and rotating maximum intensity projection (MIP) images (45 views separated by 8 degrees) were created. After rotating MIP images were randomized and anonymized, they were scored independently by two radiologists with neuroradiology fellowship training using the following 4-point scale: 1, non-diagnostic image quality; 2, fair image quality with moderate artifacts; 3, good image quality with mild artifacts; and 4, excellent image quality with minimal to no artifacts. Twenty arterial segments were scored: 1, aortic arch; 2, brachiocephalic artery; 3 and 4, bilateral common carotid arteries; 5 and 6, bilateral proximal internal carotid arteries; 7 and 8, cervical internal carotid arteries; 9 and 10, bilateral petrous internal carotid arteries; 11–18, bilateral V1, V2, V3, and V4 segments of the vertebral arteries; 19, basilar artery; and 20, the intracranial arteries. MIP rather than cross-sectional source images were used for image quality analysis as they are frequently reviewed in clinical practice for disease detection and have been used in recent research studies (12,13,19,29).

Statistical Analysis

Inter-rater agreement of image quality scores was done using quadratic-weighted Cohen’s kappa (κ) analysis; κ values for agreement were interpreted as: 0.00–0.20, slight; 0.21–0.40, fair; 0.41–0.60, moderate; 0.61–0.80, substantial; 0.81–1.00, almost perfect. Scoring data from the two radiologists were averaged on a segmental basis prior to comparison using Friedman and post-hoc Wilcoxon signed-rank tests. P-values less than 0.05 were considered statistically significant. Analyses were done in SciPy (version 1.4.1, https://scipy.org/) and R (version 3.5.2, R Foundation for Statistical Computing, Vienna) software.

RESULTS

3D tsSOS QISS and FLASH MRA displayed the carotid, vertebral, and intracranial vessels within a scan time of under 7 minutes (Figure 2). Measures of arterial SNR, and arterial-to-muscle CNR are provided in Supporting Information Table S1. The use of RMS combination of multiple TEs during 3D tsSOS QISS provided 60% and 63% increases in SNR and CNR with respect to the first TE of 1.6 ms. Figure 3 shows results obtained with 3D tsSOS QISS in the patient with severe carotid stenosis.

Figure 2.

Figure 2.

Open in a new tab

50-mm-thick sagittal maximum intensity projection images obtained (from left to right) with: 3D tsSOS QISS and FLASH MRA using root-mean-square (RMS) TE combination, 3D tsSOS QISS and 3D tsSOS FLASH using the shortest TE of 1.6 ms, scan time-matched 2D TOF and 2D QISS, and scan time-, resolution- and coverage-matched 3D TOF. Note that the best image quality is obtained with 3D tsSOS QISS using RMS reconstruction, the improved image quality obtained using RMS combination with respect to the shortest TE, the improved contrast seen with 3D QISS with respect to 3D FLASH, the improved spatial resolution and arterial display obtained with 3D tsSOS QISS with respect to 2D QISS and 2D TOF, and the improved image quality of 3D tsSOS QISS with respect to 3D TOF matched for acquisition time, spatial resolution, and anatomical coverage.

Figure 3.

Figure 3.

Open in a new tab

Comparison of thin maximum intensity projection images obtained in a patient with bilateral carotid stenosis (upper panel: mild stenosis of the left carotid internal carotid artery; lower panel: severe stenosis of the right internal carotid artery). Images from left to right were obtained with 3D tsSOS QISS with RMS TE combination, 3D tsSOS QISS acquired at the shortest TE of 1.6 ms, coverage-matched 2D TOF and 2D radial QISS, 3D TOF matching 3D tsSOS QISS for spatial resolution, coverage and scan time, limited-coverage 3D TOF acquired using clinical parameters, and computed tomographic angiography (CTA) acquired 4 months prior. Note the improved image quality of 3D tsSOS QISS RMS with respect to 2D TOF, 2D QISS, and 3D TOF, as well as the excellent agreement of 3D tsSOS QISS with respect to CTA.

When compared against 3D tsSOS FLASH, the use of inversion-recovery background suppression during 3D tsSOS QISS (see leftmost panels in Figure 2) resulted in approximately 36% and 10% increases in arterial-to-muscle CNR for the images obtained at the TE of 1.6 ms and with RMS combination of TEs, respectively. Adjusted to a voxel size of 1 mm3, SNR and CNR values for 3D tsSOS QISS with RMS combination of TEs were 3.8-fold and 3.2-fold larger than those of 2D QISS.

Image quality scores for the whole-neck ≈7-minute protocols are summarized in Figure 4; detailed statistical results are provided in Supporting Information Table S2. Figure 4a shows median image quality scores across the various arterial segments. Median and mean image quality scores (aggregating scores across all arterial segments) are summarized in Figure 4b; Friedman testing indicated the presence of a statistically significant difference in median image quality scores across techniques (P<0.001). Post-hoc Wilcoxon signed-rank tests revealed that 3D tsSOS QISS provided the best image quality of the seven techniques (P<0.05). 3D tsSOS QISS (RMS and TE1.6) and 3D tsSOS FLASH RMS provided better image quality than 2D QISS (P<0.05). All 3D tsSOS volumes except FLASH TE1.6 provided better image quality than 2D and 3D TOF (P<0.05). 3D tsSOS QISS RMS provided better image quality than 2D TOF and 3D TOF for displaying 14 of 20 and 18 of 20 arterial segments, respectively (P<0.05).

Figure 4.

Figure 4.

Open in a new tab

Image quality scores by technique. (a) Plot of median image quality by arterial segment and (b) bar graph of median and mean image quality after aggregating data across all arterial segments. Scores from the two reviewers were averaged. Lines in (a) are shown to improve readability of overlapping points and do not imply continuity. Scoring scale: 1, non-diagnostic; 2, fair; 3, good; and 4, excellent. CCA = common carotid artery; ICA = internal carotid artery.

Inter-rater agreement was substantial for 3D tsSOS QISS RMS (κ=0.67, P<0.001) and 2D TOF (κ=0.63, P<0.001), moderate for 3D tsSOS QISS TE1.6 (κ=0.57, P<0.001), 3D tsSOS FLASH (κ=0.59 for TE1.6 and κ=0.48 for RMS, P<0.001 both comparisons) and 3D TOF (κ=0.47, P<0.001), and fair for 2D QISS (κ=0.39, P<0.001).

DISCUSSION

In this work we evaluated the feasibility of using a 3D tsSOS implementation of the QISS technique at 3T to efficiently image the entire neck with improved spatial resolution in the slice direction than is possible using either 2D QISS or 2D TOF. A multi-echo FLASH readout with RMS combination of multiple TEs was used to enhance SNR. The technique portrayed the neck arteries with good to excellent image quality. Moreover, we found that 3D tsSOS QISS provided improved image quality with respect to 3D TOF, 2D QISS, and 2D TOF approaches, while providing better CNR than a comparable thin-slab SOS FLASH technique.

By using very thin 3D slabs in conjunction with a low flip angle RF excitation and QI to allow for inflow of unsaturated arterial spins, 3D tsSOS QISS provides similar benefits to 2D QISS with regard to minimizing flow saturation effects. Compared with an unmodified FLASH technique, background signal suppression was improved by the intermittent application of a slab-selective inversion pulse. By comparison, 2D and 3D TOF apply RF excitation pulses without pause, which tends to exacerbate saturation effects in the setting of recirculating flow. Moreover, the tracking venous inversion pulse is applied only once every 1.5 sec with tsSOS QISS, followed by the QI. This sequence structure reduces the risk of inadvertently saturating inferiorly directed arterial segments (e.g. in V3) compared with TOF, where the tracking venous saturation pulse is applied at short intervals of approximately 20–30 ms, with no time allowed for inflow of unsaturated arterial spins (30,31).

Some key findings of this study are that 3D tsSOS QISS provided better image quality than 2D QISS for portraying the neck arteries, approximately 2.5-fold smaller voxels, and more than 3-fold improvements in SNR and CNR when normalized for acquired voxel size. Smaller voxels are advantageous in that they improve vascular detail, facilitate more accurate measurements of arterial diameter and cross-sectional area, and reduce dephasing artifacts in the setting of accelerating or turbulent flow across a hemodynamically significant stenosis. Moreover, in comparison with 2D techniques such as 2D TOF and 2D QISS, the use of a 3D stack-of-stars acquisition helps to reduce motion artifacts from swallowing and respiration.

Another finding of our study is that the use of inversion-recovery-based background suppression with 3D tsSOS QISS improved CNR versus 3D tsSOS FLASH without the use of dedicated background signal suppression, and improved image quality at the minimum TE of 1.6 ms. The use of thin slabs ensured sufficient inflow of blood into the thin imaging slab during the QI which spanned 583 ms between the application of the inversion pulse to the start of the stack-of-stars readout. Given conservative estimates of the cycle-averaged mean velocities of 20 cm/s and 10 cm/s in the carotid and vertebral arteries respectively (32), there should be approximately 11.6 cm and 5.8 cm of inflow, which well exceeds the thin slab thickness of 2 cm.

We found that the use of a multi-echo acquisition with RMS combination of TEs provided images that were of higher quality, SNR and CNR than those obtained by reconstructing data from only the shortest TE of 1.6 ms. We attribute this to the averaging of noise-like artifacts across TEs. While RMS combination of TEs was beneficial in terms of image quality, we observed that magnetic susceptibility-related signal loss in the petrous segments of the internal carotid arteries was more pronounced in the RMS images than in those acquired at the shortest TE. Moreover, because phase offsets due to accelerating blood flow are proportional to the square of the TE, images acquired at the shortest TE are expected to display less flow-related dephasing near stenoses. Consequently, a reasonable diagnostic approach might involve review of the RMS combined images for initial whole-neck evaluation, while inspection of the shortest TE images would be helpful to minimize off-resonance artifacts near the skull base and flow-related signal dephasing through hemodynamically significant lesions. Leveraging the multi-echo readout, future work could also test the utility of Dixon- or IDEAL-based methods to generate water-only images (33–35).

The use of a multi-echo readout with tsSOS QISS prolongs the FLASH TR but in our experience is preferable over the application of shorter FLASH TRs using the same or smaller flip angles, which increase saturation artifacts and reduce SNR, respectively. We observed that a QISS TR of 1500 ms coupled with an acquisition duration of ≈890 ms and QI of ≈580 ms worked well for imaging our thin 19.5 mm thick imaging slabs. Fixing the acquisition duration and slab thickness, shorter TR and QI times improve arterial-to-background contrast (through greater inversion-recovery based signal nulling), but come at the cost of less arterial inflow into the imaging slab. Moreover, so long as spatial resolution in the slice direction and the number of acquired radial views in each slice-encoding step are fixed, the acquisition of thinner slabs reduces SNR as fewer slice-encoding steps are collected. Of note, we did not observe that intrinsic in-plane flow compensation of the second TE due to even echo rephasing improved the image quality with respect to the first and third TEs; this observation was likely due to the fact that the main direction of blood flow (foot to head) was orthogonal to the axial direction of the radial readout gradient.

3D tsSOS QISS outperformed 3D TOF MRA when the protocols were configured for whole-neck coverage and matched for acquisition time, anatomical coverage, receiver bandwidth, and spatial resolution. This may in part be due to less severe image artifacts from radial undersampling as opposed to the use of high parallel acceleration factors, which results in noise amplification for Cartesian acquisitions. More advanced acceleration strategies for 3D TOF were not available on our MRI system. Future implementations of all 3D protocols applied here could entail the integration of more advanced acceleration, reconstruction, and image enhancement strategies such as compressed sensing and deep learning (14,36,37).

CONCLUSION

In conclusion, 3D tsSOS QISS for whole-neck nonenhanced MRA is feasible at 3T, benefits from the use of a multi-echo readout with RMS combination of echo signals, provides improved resolution in the slice direction than 2D imaging methods, and in this study provided better image quality than 3D TOF when using protocols matched for scan time, anatomical coverage, and spatial resolution. Additional studies will be needed to establish the diagnostic performance of 3D tsSOS QISS in patients with known or suspected arterial disease.

Supplementary Material

1

ACKNOWLEDGEMENT

This work was supported in part by the National Institute of Biomedical Imaging and Bioengineering of the National Institutes of Health under Award Number R01EB027475. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

REFERENCES

  • 1.Levy RA, Prince MR. Arterial-phase three-dimensional contrast-enhanced MR angiography of the carotid arteries. Am. J. Roentgenol 1996;167:211–215. [DOI] [PubMed] [Google Scholar]
  • 2.Menke J Diagnostic accuracy of contrast-enhanced MR angiography in severe carotid stenosis: meta-analysis with metaregression of different techniques. Eur Radiol 2009;19:2204–2216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Kuo PH, Kanal E, Abu-Alfa AK, Cowper SE. Gadolinium-based MR contrast agents and nephrogenic systemic fibrosis. Radiology 2007;242:647–649. [DOI] [PubMed] [Google Scholar]
  • 4.Kanda T, Ishii K, Kawaguchi H, Kitajima K, Takenaka D. High signal intensity in the dentate nucleus and globus pallidus on unenhanced T1-weighted MR images: Relationship with increasing cumulative dose of a gadoliniumbased contrast material. Radiology 2014;270:834–841. [DOI] [PubMed] [Google Scholar]
  • 5.Kanda T, Fukusato T, Matsuda M, et al. Gadolinium-based contrast agent accumulates in the brain even in subjects without severe renal dysfunction: Evaluation of autopsy brain specimens with inductively coupled plasma mass spectroscopy. Radiology 2015;276:228–232. [DOI] [PubMed] [Google Scholar]
  • 6.Murata N, Gonzalez-Cuyar LF, Murata K, et al. Macrocyclic and other non-group 1 gadolinium contrast agents deposit low levels of gadolinium in brain and bone tissue: Preliminary results from 9 patients with normal renal function. Invest. Radiol 2016;51:447–453. [DOI] [PubMed] [Google Scholar]
  • 7.Nael K, Meshksar A, Ellingson B, et al. Combined low-dose contrast-enhanced MR angiography and perfusion for acute ischemic stroke at 3T: A more efficient stroke protocol. Am. J. Neuroradiol 2014;35:1078–1084. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Keller PJ, Drayer BP, Fram EK, Williams KD, Dumoulin CL, Souza SP. MR angiography with two-dimensional acquisition and three-dimensional display. Work in progress. Radiology 1989;173:527–532. [DOI] [PubMed] [Google Scholar]
  • 9.Wagle WA, Dumoulin CL, Souza SP, Cline HE. 3DFT MR angiography of carotid and basilar arteries. Am. J. Neuroradiol 1989;10:911–919. [PMC free article] [PubMed] [Google Scholar]
  • 10.Parker DL, Yuan C, Blatter DD. MR angiography by multiple thin slab 3D acquisition. Magn Reson Med 1991;17:434–451. [DOI] [PubMed] [Google Scholar]
  • 11.Blatter DD, Bahr AL, Parker DL, et al. Cervical carotid MR angiography with multiple overlapping thin-slab acquisition: comparison with conventional angiography. AJR Am J Roentgenol 1993;161:1269–1277. [DOI] [PubMed] [Google Scholar]
  • 12.Koktzoglou I, Aherne EA, Walker MT, Meyer JR, Edelman RR. Ungated nonenhanced radial quiescent interval slice-selective (QISS) magnetic resonance angiography of the neck: Evaluation of image quality. J. Magn. Reson. Imaging 2019;50:1798–1807. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Peters S, Huhndorf M, Jensen-Kondering U, et al. Non-contrast-enhanced carotid MRA: Clinical evaluation of a novel ungated radial quiescent-interval slice-selective MRA at 1.5T. Am. J. Neuroradiol 2019;40:1529–1537. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Koktzoglou I, Huang R, Ong AL, Aouad PJ, Aherne EA, Edelman RR. Feasibility of a sub-3-minute imaging strategy for ungated quiescent interval slice-selective MRA of the extracranial carotid arteries using radial k-space sampling and deep learning–based image processing. Magn. Reson. Med 2020;84:825–837. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Edelman RR, Koktzoglou I. Noncontrast MR angiography: An update. J. Magn. Reson. Imaging 2019;49:355–373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Edelman RR, Aherne E, Leloudas N, Pang J, Koktzoglou I. Near-isotropic noncontrast MRA of the renal and peripheral arteries using a thin-slab stack-of-stars quiescent interval slice-selective acquisition. Magn. Reson. Med 2020;83:1711–1720. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Edelman RR, Sheehan JJ, Dunkle E, Schindler N, Carr J, Koktzoglou I. Quiescent-interval single-shot unenhanced magnetic resonance angiography of peripheral vascular disease: Technical considerations and clinical feasibility. Magn. Reson. Med 2010;63:951–958. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Edelman RR, Carr M, Koktzoglou I. Advances in non-contrast quiescent-interval slice-selective (QISS) magnetic resonance angiography. Clin. Radiol 2019;74:29–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Koktzoglou I, Murphy IG, Giri S, Edelman RR. Quiescent interval low angle shot magnetic resonance angiography of the extracranial carotid arteries. Magn. Reson. Med 2016;75:2072–2077. [DOI] [PubMed] [Google Scholar]
  • 20.Koktzoglou I, Edelman RR. Super-resolution intracranial quiescent interval slice-selective magnetic resonance angiography. Magn. Reson. Med 2018;79:683–691. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Serhal A, Koktzoglou I, Aouad P, et al. Cardiovascular magnetic resonance imaging of aorto-iliac and ilio-femoral vascular calcifications using proton density-weighted in-phase stack of stars. J. Cardiovasc. Magn. Reson 2018;20:51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Carr HY. Steady-state free precession in nuclear magnetic resonance. Phys. Rev 1958;112:1693–1701. [Google Scholar]
  • 23.Koktzoglou I, Edelman RR. Radial fast interrupted steady-state (FISS) magnetic resonance imaging. Magn. Reson. Med 2018;79:2077–2086. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Shonai T, Carpenter JS, Lemieux SK, et al. Improvement of vessel visibility in time-of-flight MR angiography of the brain. J. Magn. Reson. Imaging 2008;27:1362–1370. [DOI] [PubMed] [Google Scholar]
  • 25.Nitz WR. Fast and ultrafast non-echo-planar MR imaging techniques. Eur. Radiol 2002;12:2866–2882. [DOI] [PubMed] [Google Scholar]
  • 26.Martin N, Malfair D, Zhao Y, et al. Comparison of MERGE and axial T2-weighted fast spin-echo sequences for detection of multiple sclerosis lesions in the cervical spinal cord. Am. J. Roentgenol 2012;199:157–162. [DOI] [PubMed] [Google Scholar]
  • 27.Higashihira S, Kobayashi N, Oishi T, et al. Comparison Between 3-Dimensional Multiple-Echo Recombined Gradient Echo Magnetic Resonance Imaging and Arthroscopic Findings for the Evaluation of Acetabular Labrum Tear. Arthrosc. - J. Arthrosc. Relat. Surg 2019;35:2857–2865. [DOI] [PubMed] [Google Scholar]
  • 28.Firbank MJ, Coulthard A, Harrison RM, Williams ED. A comparison of two methods for measuring the signal to noise ratio on MR images. Phys Med Biol 1999;44:N261–4. [DOI] [PubMed] [Google Scholar]
  • 29.Koktzoglou I, Walker MT, Meyer JR, Murphy IG, Edelman RR. Nonenhanced hybridized arterial spin labeled magnetic resonance angiography of the extracranial carotid arteries using a fast low angle shot readout at 3 Tesla. J. Cardiovasc. Magn. Reson 2016;18:18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Wikstrom J, Bjornerud A, McGill S, Johansson L. Venous saturation slab causes overestimation of stenosis length in two-dimensional time-of-flight magnetic resonance angiography. Acta radiol. 2009. doi: 10.1080/02841850802562063. [DOI] [PubMed] [Google Scholar]
  • 31.Lim RP, Koktzoglou I. Noncontrast Magnetic Resonance Angiography: Concepts and Clinical Applications. Radiol. Clin. North Am 2015;53:457–476. [DOI] [PubMed] [Google Scholar]
  • 32.Yazici B, Erdoğmuş B, Tugay A. Cerebral blood flow measurements of the extracranial carotid and vertebral arteries with Doppler ultrasonography in healthy adults. Diagn. Interv. Radiol 2005;11:195–8. [PubMed] [Google Scholar]
  • 33.Dixon WT. Simple proton spectroscopic imaging. Radiology 1984;153:189–194. [DOI] [PubMed] [Google Scholar]
  • 34.Reeder SB, McKenzie CA, Pineda AR, et al. Water-fat separation with IDEAL gradient-echo imaging. J. Magn. Reson. Imaging 2007;25:644–652. [DOI] [PubMed] [Google Scholar]
  • 35.Grayev A, Shimakawa A, Cousins J, Turski P, Brittain J, Reeder S. Improved time-of-flight magnetic resonance angiography with IDEAL water-fat separation. J. Magn. Reson. Imaging 2009;29:1367–1374. [DOI] [PubMed] [Google Scholar]
  • 36.Fushimi Y, Fujimoto K, Okada T, et al. Compressed Sensing 3-Dimensional Time-of-Flight Magnetic Resonance Angiography for Cerebral Aneurysms: Optimization and Evaluation. Invest. Radiol 2016;51:228–235. [DOI] [PubMed] [Google Scholar]
  • 37.Li B, Li H, Dong L, Huang G. Fast carotid artery MR angiography with compressed sensing based three-dimensional time-of-flight sequence. Magn. Reson. Imaging 2017;43:129–135. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1627958-supplement-1.pdf (106.1KB, pdf)

RESOURCES