Continuously moving table aorto-iliofemoral run-off contrast-enhanced magnetic resonance angiography: image quality analysis in comparison to the multistep acquisition

Abstract

Background

Optimal vessel contrast is a prerequisite for vascular imaging. Consecutive stationary imaging of multiple fields of view is contrary to the continuous contrast material passage through the vascular tree. A continuous acquisition of a magnetic resonance (MR) sequence might overcome this limitation.

Purpose

To investigate the image quality of a continuously moving table (CMT) acquisition compared with the established multistep approach for contrast-enhanced magnetic resonance angiography (CE-MRA) of the aorto-iliofemoral run-off.

Material and Methods

Institutional review board approved this retrospective interindividual study of 60 consecutive patients referred to CE-MRA for peripheral arterial disease. Thirty patients underwent CE-MRA using the routine multistep acquisition and 30 patients were scanned using the CMT technique at 1.5 Tesla. All patients received a fixed contrast dose of 25 mL gadoterate meglumine. A quantitative analysis was performed to assess the relative contrast of 10 vascular segments from the proximal abdominal aorta to the distal calf arteries. A qualitative evaluation of three separate vascular regions (abdomen and pelvis, thighs, and calves) was performed. Two radiologists graded independently arterial vessel conspicuity, venous contamination, presence of artifacts, and diagnostic confidence on a 4-point scale. Overall scan time, including all localizer scans, was recorded. Statistical differences were tested using the Wilcoxon signed-rank test with Bonferroni correction.

Results

No significant differences were found between the continuously moving table acquisition and the multistep acquisition with regard to the relative vascular contrast and the qualitative image criteria. The agreement between both readers was significant (Kendall tau rank correlation coefficient, 0.373). The absolute reader agreement was 71.4%. The mean overall scan time was 12 min 44 s for the CMT protocol and 21 min 41 s for the multistep protocol.

Conclusion

Aorto-iliofemoral run-off CE-MRA acquired with CMT technique provides a high image quality equivalent to a multistep technique at an overall scan time reduction of 41.3%.

Keywords

Adults arteries arteriosclerosis MR angiography vascular

Introduction

Peripheral arterial disease (PAD) is a common manifestation of atherosclerotic disease, affecting 12–14% of the general population and as many as 20% of individuals over the age of 75 years (1). Computed tomography (CT) is currently the most frequently used non-invasive vascular imaging modality to diagnose PAD, since it has proven not only to have excellent diagnostic performance but also to be effective in guiding therapeutic decision-making in PAD patients (2). The major drawbacks of CT angiography are beam hardening artifacts related to calcified plaques and nephrotoxicity of the iodine-based contrast medium, especially since impaired kidney function is a common co-morbidity in PAD patients (3).

Technical improvements in gradient performance and coil technology have improved the capability of imaging the vascular tree by magnetic resonance imaging (MRI) using high spatial resolution gradient-echo sequences. The inherent advantage of MRI over CT is the absence of artifacts related to vessel wall calcifications resulting in the inability to assess the degree of stenosis (4). From a practical clinical perspective, the relatively fast and easy generation of maximum intensity projection reconstructions has resulted in frequent vascular imaging by MRI. However, optimal vessel contrast without venous contamination is not always achievable especially when scanning a large z-axis field of view (FOV) (e.g. peripheral run-off). This is likely related to consecutive imaging of multiple fields of view, which is conflicting with the continuous contrast bolus passage through the vascular system after intravenous bolus injection. A continuous acquisition of the MR sequence might overcome this limitation by more optimally linking contrast bolus passage and scanning, similar to multislice CT, potentially improving image quality. Improvements in table hardware, sequence design, and reconstruction software allow the acquisition of MR images during movement of the patient table (5 –8). A previous study has shown that contrast-enhanced (CE) continuously moving table MR angiography (CMT-MRA) provides robust image quality at a reduced overall scan time (9). However, a dedicated head-to-head comparison between a multistep protocol and a continuously moving table protocol with identical scanner hardware and contrast medium application on a larger patient cohort has not been published. Therefore, a CMT protocol was implemented for CE-MRA of the aorto-iliofemoral run-off. The purpose of this study was to investigate the image quality of a CMT acquisition CE-MRA compared to the established multistep approach for CE-MRA of the aorto-iliofemoral run-off.

Material and Methods

Patients

Sixty consecutive patients, referred from a vascular surgery outpatient clinic with high clinical suspicion for or known aorto-iliofemoral vascular disease, were included in this interindividual study and evaluated retrospectively. Thirty consecutive patients (18 men, 12 women; mean age, 62.7 ± 11.5 years; mean body weight, 82.4 ± 16.9 kg; mean height, 174 ± 10 cm; mean body surface area, 1.99 m²) underwent CE-MRA using a routine multistep acquisition during a 7-month period. Thirty additional patients (22 men, 8 women; mean age, 58.9 ± 12.2 years; mean body weight, 80.8 ± 18.1 kg; mean height, 172 ± 9 cm; mean body surface area, 1.96 m²), serving as a comparison cohort, were subsequently scanned using the CE-MRA protocol with CMT technique during a 5-month interval. This interindividual study design with two consecutive cohorts was chosen, since an intraindividual comparison study was discouraged by the institutional review board (IRB) due to concerns regarding repetitive contrast medium application. The interindividual study was approved by the local institutional review board, and written informed consent was waived for retrospective image and data analysis.

MRI and contrast medium application

MRI was performed on a 1.5 T system with 32 independent RF receiver channels using phased-array surface coils and an array of spine coils integrated in the patient table (Magnetom Avanto with total imaging matrix (TIM) coil technology, Software Version VB15, Siemens Healthcare Sector, Erlangen, Germany). The localizer scans were obtained either as one CMT scan (in three planes) or as three separate scans at different table positions (in three planes each). Phase-contrast vessel scouts were obtained at three FOVs followed by a test-bolus acquisition at the level of the origin of the renal arteries. Subsequently, a three-dimensional (3D) gradient-echo sequence (fast low-angle shot [FLASH]) was acquired in the coronal orientation before and after contrast medium administration. The sequence protocols, therefore, comprised seven sequences in the CMT group and 13 sequences in the multistep group. All subjects were imaged feet first and were asked to hold their breath at the beginning of the acquisition as long as possible. The sequence parameters for the CMT and multistep (values in parenthesis) MRA sequences, respectively, were as follows: TR, 2.4 ms (3.2 ms); TE, 0.8 ms (1.2 ms); bandwidth, 815 Hz/pixel (420 Hz/pixel); slice thickness, 1.3 mm (1.3 mm); 88 images (three sequences of 88 images); FOV, 344 × 1281 mm (375 × 500 mm with 100 mm overlap between each FOV), parallel imaging with an acceleration factor of 2 using the generalized autocalibrating partially parallel acquisitions (GRAPPA) algorithm with 24 reference lines (both protocols), flip angle 25° (25°), acquisition time 63 s (three separate sequences of 18 s acquisition time each separated by two table movements of 3 s each). The table speed of the CMT protocol was 24 mm/s and could not be adjusted individually. In the CMT protocol, an extended longitudinal FOV was acquired in one seamless image by continuous acquisition of MR data as the patient table moved through the region of interest. Data were acquired from a homogeneous central volume of the magnet, and an extension of the field of view in the z-direction was realized by spatially registering the data (7).

The CMT acquisition and the multistep approach were acquired after antecubital bolus administration of 25 mL gadoterate meglumine contrast medium (0.5 mmol gadolinium/mL Dotarem®, Guerbet, Roissy, France). A two-phase contrast medium injection protocol was used, with the initial 15 mL injected at a flow rate of 1 mL/s and the last 10 mL injected at a flow rate of 0.5 mL/s. The contrast medium was followed by a saline flush of 15 mL at a flow rate of 0.5 mL/s. The patient’s individual circulation time was calculated by an axial test-bolus sequence at the level of the renal arteries using 2 mL of contrast medium with the region of interest placed in the aorta. Three seconds were added to the peak of the circulation time before starting the MRA sequence. The body weight-adapted average total contrast dose per patient cohort was calculated. No venous compression techniques were used.

Quantitative evaluation

The quantitative analysis measured the relative contrast of 10 vessel segments (aorta at the level of the origin of the right renal artery, aortic bifurcation, right common iliac artery, right common femoral artery, mid-segment of the right superficial femoral artery, right popliteal artery, proximal and distal right anterior and posterior tibial arteries) with respect to a reference muscle, which was chosen as closely located to the vessel as anatomically possible to minimize influence of the different coil sensitivity profiles. In case of high-grade stenosis of the right lower extremity and absence of high-grade stenosis of the left lower extremity, the corresponding left lower extremity arteries were chosen instead. Vascular contrast was calculated as the relative signal intensity (rSI) according to the formula

rSI = \frac{SIvessel - SIreferencemuscle}{SIvessel + SIreferencemuscle}

for each respective vessel. Occluded vessels without distal collateral filling were excluded from the analysis. The contrast-to-noise ratio (CNR) was not measured due to the application of parallel imaging, which implies a spatially inhomogeneous noise distribution (10). The overall scan time (including the time needed for planning the sequences) of both sequence protocols (CMT and multistep) including all localizer and test-bolus acquisitions were extracted from the Digital Imaging and Communications in Medicine (DICOM) headers and recorded.

Qualitative evaluation

The qualitative evaluation was performed independently by two radiologists (JOH, RJS), with 14 and 18 years of experience in vascular MRI, respectively, who were blinded to the acquisition technique and any clinical information. Also, the readers were unaware of the fact that a continuous moving table technique was used, as the images were presented to them on a picture archiving and communication system (PACS) workstation with the CMT-MRA stacks zoomed to the size of the respective vascular region such that the nature of the one seamless image stack was not evident. On a 4-point scale, the two readers graded arterial vessel conspicuity (4, excellent; 3, good; 2, moderate; 1, inadequate), venous contamination (4, no; 3, mild; 2, moderate; 1, severe), presence of artifacts (4, no; 3, mild; 2, moderate; 1, severe), and diagnostic confidence (4, definite diagnosis achievable; 3, probable diagnosis achievable; 2, uncertain diagnosis; 1, non-diagnostic) for three separate vascular regions (abdomen and pelvis, thighs, and calves).

Statistical analysis

Statistical calculations were performed using the SPSS software (version 11.5; SPSS Inc., Chicago, IL, USA). The non-parametric Wilcoxon signed-rank test was used and a P value < 0.05 was considered to indicate statistical significance. A Bonferroni correction for multiple comparisons was carried out, resulting in corrected P values of 0.005 for the quantitative analysis and 0.004 for the qualitative analysis. Reader agreement was assessed by calculation of the Kendall tau rank correlation coefficient and by the absolute reader agreement.

Results

The average body weight-adapted total contrast dose was 0.155 ± 0.035 mmol/kg for the CMT-MRA cohort versus 0.152 ± 0.039 mmol/kg for the multistep MRA cohort.

Due to vessel occlusion without distal collateral filling, 10 vessel segments from the CMT-MRA cohort and eight segments from the multistep cohort were excluded from the analysis.

No significant differences were found between the two imaging protocols regarding the relative vascular signal intensities (Table 1).

Table 1.

Quantitative analysis of the relative vascular contrast.

Vessel segment	Relative signal intensity (rSI)		P value*
Vessel segment	CMT	Multistep	P value*
Aorta at the level of the origin of the right renal artery	0.752 ± 0.098	0.781 ± 0.087	0.360
Aortic bifurcation	0.804 ± 0.066	0.796 ± 0.090	0.719
Right common iliac artery	0.781 ± 0.099	0.804 ± 0.092	0.600
Right common femoral artery	0.694 ± 0.145	0.705 ± 0.010	0.992
Middle segment of the right superficial femoral artery	0.748 ± 0.098	0.730 ± 0.167	0.873
Right popliteal artery	0.724 ± 0.075	0.715 ± 0.076	0.905
Proximal right anterior tibial artery	0.683 ± 0.108	0.698 ± 0.081	0.517
Proximal right posterior tibial artery	0.634 ± 0.144	0.648 ± 0.120	0.551
Distal right anterior tibial artery	0.463 ± 0.132	0.539 ± 0.109	0.007
Distal right posterior tibial artery	0.462 ± 0.120	0.523 ± 0.150	0.171

The relative signal intensities are given as mean ± standard deviation.

Adjusted level of statistical significance 0.005 (Bonferroni correction).

CMT, continuously moving table.

The mean overall scan time was 41.3% less for the CMT protocol (12 min 44 s ± 51 s; mean ± standard deviation) compared to the multistep protocol (21 min 41 s ± 1 min 37 s; mean ± standard deviation).

In the qualitative evaluation, no significant differences between the two MRA protocols were found regarding arterial vessel conspicuity, venous contamination, artifacts, and diagnostic confidence (Table 2). The agreement between both readers was significant with a Kendall tau rank correlation coefficient of 0.373. Additionally, the absolute reader agreement was 71.4%. Representative maximum-intensity-projection reconstruction images of both acquisition techniques are given in Figs. 1 –3.

Fig. 1.

CMT-MRA, maximum intensity projection reconstruction in a 71-year-old male patient with infrarenal abdominal aortic aneurysm and right popliteal artery aneurysm treated by venous bypass graft (aneurysm is not completely cut off from blood flow). Some venous contamination is seen in the legs.

Fig. 2.

CMT-MRA, maximum intensity projection reconstruction in a 61-year-old female patient with high-grade stenosis at the origin of the left common iliac artery and complete occlusion of the left superficial femoral artery with collateral run-off to the leg via the deep femoral artery. Venous contamination is seen primarily in the legs and feet.

Fig. 3.

Multistep MRA, maximum intensity projection reconstruction composed from three steps in a 63-year-old male patient with suspected peripheral artery occlusive disease and a high-grade stenosis of the distal right superficial femoral artery. There is minimal venous contamination.

Table 2.

Qualitative analysis of the image quality criteria and diagnostic confidence.

	Abdomen and pelvis			Thighs			Calves
	CMT	Multistep	P value	CMT	Multistep	P value	CMT	Multistep	P value
Arterial conspicuity	3.7	3.8	n.s.	3.6	3.8	n.s.	3.3	3.5	n.s.
Venous contamination	3.6	3.6	n.s.	3.2	3.2	n.s.	2.6	2.8	n.s.
Presence of artifacts	3.4	3.2	n.s.	3.4	3.3	n.s.	3.3	3.4	n.s.
Diagnostic confidence	3.7	3.9	n.s.	3.7	3.7	n.s.	3.2	3.4	n.s.

All values are given as means. Arterial conspicuity (4, excellent; 3, good; 2, moderate; 1, inadequate); venous contamination (4, no; 3, mild; 2, moderate; 1, severe); presence of artifacts (4, no; 3, mild; 2, moderate; 1, severe); diagnostic confidence (4, definite diagnosis achievable; 3, probable diagnosis achievable; 2, uncertain diagnosis; 1, non-diagnostic).

CMT, continuously moving table; n.s., not statistically significant.

Discussion

CE-MRA of the aorto-iliofemoral run-off has proven to be a reliable modality for assessing peripheral arterial disease and has become the preferred imaging technique in many institutions (4). A multistep acquisition at different table positions (usually three or four) is most commonly used and is referred to as the bolus-chase technique. Since a large z-axis field of view needs to be imaged in aorto-iliofemoral runoff MRA, it is critical to acquire all images during the arterial phase in order to avoid venous contamination. Also, image quality may be decreased due to image distortion at the edges of each field of view as a consequence of gradient non-linearities (11). The concept of a continuously moving table acquisition targets these limitations.

The feasibility of CMT has been investigated in previous studies on volunteers and a small number of patients (5,6,12). A study by Koziel et al. on 82 patients has shown that CMT-MRA provides a robust image quality (8). However, a dedicated comparison between a multistep protocol and a continuously moving table protocol with identical scanner hardware and contrast medium application on a larger patient cohort with regard to vessel conspicuity and image quality has not been published. This comparative image quality analysis on a cohort of 60 patients shows that CMT-MRA can provide the same level of image quality as the multistep protocol, however, no increase in image quality could be observed with CMT-MRA; arterial vessel conspicuity and diagnostic confidence were rated on an equal level. The quantitative evaluation of vascular contrast matched with the qualitative evaluation of arterial vessel conspicuity, showing no significant differences between both acquisition techniques and a tendency for decreasing vessel conspicuity and vascular contrast from the proximal to the distal imaging FOV. The most distal segments were still rated diagnostic with overall good vessel conspicuity on both imaging protocols. In addition, the same level of mild to moderate venous contamination was present in both acquisition protocols, which is in contrast to a previous study that reported lesser venous contamination using a CMT-MRA protocol compared to a multistep protocol (8). These differences might be explained by a different contrast medium injection protocol since we used a lower flow rate compared to Koziel et al. Different table speeds might also improve the adjustment of contrast bolus passage to scanning. In the scanner set-up that was used in this study, the table speed could not be modified. Parallel imaging in multiple directions might also be used in the future to decrease scan time in order to facilitate a more optimal bolus timing (13). Overall, the CMT-MRA protocol could not improve the image quality compared to the multistep acquisition, although CMT-MRA allows scanning in the isocenter, resulting in optimal magnetic field homogeneity and gradient linearity. We therefore speculate that the contrast medium injection protocol is the most crucial point and needs to be further optimized.

A distinct advantage of CMT-MRA is improved workflow, since planning, acquisition and postprocessing of several FOVs in the multistep approach is complex, time-consuming, and susceptible to planning errors. In the multistep MRA protocol used in this study a total of 13 sequences were acquired. In contrast, only one large image stack needs to be positioned for CMT imaging, the number of sequences could be consecutively reduced to seven for the CMT-MRA protocol. In this study, we could show a reduction of the overall scan time of 41.3%, which is nearly in concordance with Koziel et al., where the overall scan time reduction was 34% compared to the multistep approach (9). However, only one image stack is used for CMT-MRA and all vascular territories should be covered in one single coronal stack, which ultimately may result in more slices and consecutively more acquisition time needed for the MRA sequence compared to the multistep approach, where the number of slices per stack can be individually chosen and angulation for each vascular territory can be performed separately.

This retrospective study has several limitations. The interindividual study design might cause bias due to differences in both cohorts. In addition, we used a fixed contrast medium volume to simplify the two-phase contrast injection, which might result in bias especially in patients with high body weight. However, the average body weight-adapted total contrast dose and patient age were nearly identical in both cohorts. Finally, no evaluation of diagnostic accuracy for assessing stenosis has been performed, since this was not the focus of this comparative study of two CE-MRA techniques focusing on image quality.

In conclusion, aorto-iliofemoral run-off contrast-enhanced MR angiography using a CMT technique is comparable to a multistep technique in image quality, while it saves acquisition time.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Hiatt

Hoag

Hamman

. Arterial disease/hypertension/angiotensin system: effect of diagnostic criteria on the prevalence of peripheral arterial disease: the San Luis Valley diabetes study. Circulation 1995; 91: 1472–1479.

Napoli

Anzidei

Zaccagna

. Peripheral arterial occlusive disease: diagnostic performance and effect on therapeutic management of 64-section CT angiography. Radiology 2011; 261: 976–986.

De Graauw

Chonchol

Poppert

. Relationship between kidney function and risk of asymptomatic peripheral arterial disease in elderly subjects. Nephrol Dial Transplant 2011; 26: 927–932.

Norgren

Hiatt

Dormandy

. On behalf of the TASC II working group. Inter-society consensus for the management of peripheral arterial disease (TASC II). J Vasc Surg 2007; 45: S5A–67A.

Madhuranthakam

Kruger

Riederer

. Time-resolved 3D contrast-enhanced MRA of an extended FOV using continuous table motion. Magn Reson Med 2004; 51: 568–576.

Kruger

Riederer

Rossman

. Recovery of phase inconsistencies in continuously moving table extended field of view magnetic resonance imaging acquisitions. Magn Reson Med 2005; 54: 712–717.

Kruger

Riederer

Grimm

. Continuously moving table data acquisition method for long FOV contrast-enhanced MRA and whole-body MRI. Magn Reson Med 2002; 47: 224–231.

Zenge

Vogt

Brauck

. High-resolution continuously acquired peripheral MR angiography featuring partial parallel imaging GRAPPA. Magn Reson Med 2006; 56: 859–865.

Koziel

Attenberger

Lederle

. Peripheral MRA with continuous table movement: imaging speed and robustness compared to a conventional stepping table technique. Eur J Radiol 2011; 80: 537–542.

10.

Dietrich

Raya

Reeder

. Measurement of signal-to-noise ratios in MR images: influence of multichannel coils, parallel imaging, and reconstruction filters. J Magn Reson Imaging 2007; 26: 375–385.

11.

Vogt

Zenge

Ladd

. Peripheral vascular disease: comparison of continuous MR angiography and conventional MR angiography: pilot study. Radiology 2007; 243: 229–238.

12.

Kramer

Zenge

Schmitt

. Peripheral magnetic resonance angiography (MRA) with continuous table movement at 3.0 T: initial experience compared with step-by-step MRA. Invest Radiol 2008; 43: 627–634.

13.

Kukuk

Hadizadeh

Gieseke

. Highly undersampled supraaortic MRA at 3.0 T: initial results with parallel imaging in two directions using a 16-channel neurovascular coil and parallel imaging factors up to 16. Magn Reson Imaging 2010; 28: 1311–1318.