Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

DOI: 10.1148/radiol.2313021113
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chen, Q.
Right arrow Articles by Edelman, R. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, Q.
Right arrow Articles by Edelman, R. R.
(Radiology 2004;231:893-899.)
© RSNA, 2004

Technical Developments

On Improving Temporal and Spatial Resolution of 3D Contrast-enhanced Body MR Angiography with Parallel Imaging1

Qun Chen, PhD, Carla V. Quijano, MD, Vu M. Mai, PhD, Saravanan K. Krishnamoorthy, MD, Wei Li, MD, Pippa Storey, PhD and Robert R. Edelman, MD

1 From the Department of Radiology, Evanston Northwestern Healthcare, Evanston, Ill (Q.C., C.V.Q., V.M.M., W.L., P.S., R.R.E.); and Department of Radiology, Northwestern University Feinberg School of Medicine, Chicago, Ill (Q.C., V.M.M., S.K.K., W.L., P.S., R.R.E.). Received September 3, 2002; revision requested November 4; final revision received September 22, 2004; accepted October 20. Supported in part by a research grant from the National Institutes of Health (RRE, R01 HL 63690–01). Address correspondence to Q.C., Department of Radiology, New York University School of Medicine, 650 First Ave, Room 600A, New York, NY 10016 (e-mail: qun.chen@med.nyu.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results
 Discussion
 REFERENCES
 
Use of a parallel imaging technique to improve temporal and spatial resolution at three-dimensional contrast-enhanced magnetic resonance (MR) angiography was investigated. Thirty experiments were performed in five groups of healthy subjects. In groups 1–3, the technique was used to improve imaging speed by a factor of two or four while maintaining spatial resolution. Contrast agent concentration was two to four times higher than at standard MR angiography, to take advantage of the faster imaging speed. In groups 4 and 5, the technique was used to double spatial resolution in the phase-encoding direction while maintaining imaging speed and contrast agent concentration. At a two to four times faster imaging speed, signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) almost equaled those at standard MR angiography, likely a result of increased contrast agent concentration. The use of parallel imaging to achieve higher spatial resolution was also proved feasible, but with substantial reduction in SNR and CNR.

© RSNA, 2004

Index terms: Magnetic resonance (MR), high-resolution • Magnetic resonance (MR), technology • Magnetic resonance (MR), three-dimensional • Magnetic resonance (MR), vascular studies


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results
 Discussion
 REFERENCES
 
Three-dimensional (3D) contrast material–enhanced body magnetic resonance (MR) angiography is a useful diagnostic technique that has become widely accepted in clinical settings (1–9). Ideally, a short imaging time is required in order to permit breath holding and to maintain preferential arterial enhancement during the first pass of the gadolinium chelate. However, the speed and spatial resolution of current imaging techniques are determined and ultimately limited by the performance characteristics of the gradient hardware. For example, standard 3D contrast-enhanced MR angiography requires a breath hold of approximately 20 seconds to acquire a 3D volume data set, and this acquisition must be performed during the first pass of the contrast material bolus, before the onset of venous enhancement. This method, however, has two drawbacks.

The first drawback is the long imaging time. The need to perform 20-second breath holds may not be tolerated by some patients, especially those with respiratory disorders. Moreover, the long imaging time also makes it crucial to synchronize the 3D MR angiographic data acquisition with the arrival of the contrast agent. A mismatch between the peak contrast enhancement and the central lines of k space would greatly degrade the quality of acquired images. If data acquisition were started too late, for example, venous enhancement would cause the interpretation of data to be difficult or impossible (3,9–11).

The second drawback is low spatial resolution: 3D contrast-enhanced MR angiography offers suboptimal spatial resolution because of restrictions caused by the total imaging time, the repetition time, and the imaging volume. While shortening the repetition time is a solution to the problem, there is a limit to how much it can be shortened. Given that the gradient performance is restricted by electrical current requirements, and, more importantly, by the potential for neuromuscular stimulation from rapid gradient switching, it is not realistic to substantially reduce repetition time and thus increase spatial resolution much beyond what is currently achieved.

With the recent introduction of parallel imaging techniques, it has become technically possible to speed up MR data acquisition without increasing gradient demands (12–20). Parallel imaging techniques such as simultaneous acquisition of spatial harmonics (12) or sensitivity encoding (15) use spatial information contained in the component elements of a coil array to partially replace spatial encoding, which would normally be performed by using gradients, thereby reducing imaging time and/or increasing spatial resolution (12,14–16,21–26).

The purpose of our study was to investigate the strengths and weaknesses of a parallel imaging technique to improve the temporal and spatial resolution at 3D contrast-enhanced MR angiography.


   Materials and Methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
 Discussion
 REFERENCES
 
Five groups, each with six healthy subjects, were examined in a total of 30 experiments. In each subject, two MR angiography examinations, one with standard technique and one with array spatial sensitivity encoding technique (ASSET), were performed during a single experiment. Twenty-two healthy subjects were examined; there were 17 men (age range, 20–48 years; mean age, 30.4 years ± 8.7 [SD]) and five women (age range, 29–56 years; mean age, 42.6 years ± 10.2). Some subjects participated in more than one group. The health status of subjects was assessed during an interview prior to the start of the study to determine whether they had a history of cardiovascular disease. Only those with no history of cardiovascular disease were included in the study. Subjects were randomly assigned to a group, and no subject was studied more than once within a specific group. For each experiment, two 3D contrast-enhanced MR angiographic data sets were obtained in random order; one data set was obtained with the standard technique, and the other was obtained with ASSET. This randomization for selecting whether ASSET or standard MR angiography would be performed first was achieved by keeping the operators (Q.C., W.L.) blinded from knowing how other experiments within the same group were conducted. The two sets of data were obtained 30 minutes apart to allow for washout of the contrast agent. The study protocol was approved by the hospital’s institutional review board, and informed consent was obtained from all subjects prior to the experiments.

Twenty-four experiments were performed by using a 1.5-T MR imager (Signa CVi; GE Medical Systems, Milwaukee, Wis) with a standard torso four-element coil array. The four elements of the coil array were placed around the subject’s abdomen, in the left-to-right direction. Two elements were placed above and two beneath the subject. Six experiments were performed on a different 1.5-T MR imager (Signa Excite Lx; GE Medical Systems) to achieve an acceleration of four times the imaging speed. This imager is equipped with eight receiver channels, each with a 1-MHz sampling bandwidth. A specially built eight-element coil array (Nova Medical, Wakefield, Mass) was used for these six experiments. The eight elements of the coil array were placed around the subject’s abdomen in the left-to-right direction. Four elements were placed above and four were placed beneath the subject.

ASSET MR Angiography
ASSET is a parallel imaging technique developed by GE Medical Systems that is implemented on the two 1.5-T MR imagers used in this study. The technique is the same as the sensitivity encoding technique (15), in which the spatial information related to the spatially varying sensitivity of receiver coils is used to reduce the number of k-space lines needed to form an image. In brief, reduced fields of view were used either to increase imaging speed while maintaining spatial resolution or to increase spatial resolution while maintaining imaging speed. Image data acquired with these reduced fields of view were then reconstructed by using the ASSET reconstruction algorithm. Reference images were acquired at the beginning of each ASSET examination to obtain the sensitivity information of the receiver coils.

3D Contrast-enhanced MR Angiography
The 3D contrast-enhanced MR angiographic data were obtained by using a fast spoiled gradient-echo sequence after administration of gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Montville, NJ), which is a T1-shortening contrast agent. A power injector (Spectris MR injection system; Medrad, Indianola, Pa) was used for administration of the contrast agent. The subjects were placed in the supine position. A volume centered on the abdominal aorta at the level of the renal arteries was chosen for the 3D data acquisition in an oblique coronal plane (Fig 1). The in-plane phase-encoding direction was left-to-right. In each subject, the same contrast agent dose (0.10 or 0.15 mmol per kilogram of body weight) was used at both ASSET MR angiography and standard MR angiography, which were performed in random order. The duration of administration of the contrast agent bolus was set to be the same as the total 3D imaging time. This resulted in the following contrast agent injection rate for all studies: injection rate = (BW x CD)/IT, where BW is the patient’s body weight, CD is the contrast agent dose, and IT is the total imaging time. Imaging parameters were 5.1/1.3 (repetition time msec/echo time msec), flip angle of 40°, field of view of 35–40 cm, and bandwidth of 125 kHz. These parameters are similar to those used in standard clinical practice at our hospital. Thirty-two 3-mm-thick sections were obtained and then interpolated to 64 sections. Eight outer sections were discarded because of imperfect section profile, resulting in 56 sections of 1.5-mm nominal section thickness. Data were acquired during breath holds. A test bolus of 1 mL of gadopentetate dimeglumine was used before each experiment to determine the arrival time of the contrast agent. A two-dimensional fast gradient-echo imaging sequence with inversion recovery and 2-second temporal resolution was used for this purpose.



View larger version (168K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Sagittal scout image with a fast spoiled gradient-echo MR sequence (15/3.5, flip angle of 20°, left lateral view) shows the volume selected for 3D data acquisition in an oblique coronal plane.

 
In the study of increased imaging speed, subjects were divided into three groups (groups 1–3, Table 1). In this study, the matrix size of the reconstructed images was kept the same while the imaging time was reduced from 20 seconds to 10 and 5 seconds by using ASSET. Since the same amount of contrast agent was used at both ASSET and standard MR angiography, the contrast agent injection rate for ASSET MR angiography in these groups, accordingly, was two to four times faster to take advantage of the reduced imaging time. This faster injection rate resulted in a two to four times higher contrast agent concentration compared with that at standard MR angiography.


View this table:
[in this window]
[in a new window]

  		TABLE 1. ASSET MR Angiography with Acceleration Factors of Two and Four to Increase Imaging Speed

 
In the study of increased spatial resolution, subjects were divided into two groups that received different doses of the contrast agent (Table 2, groups 4 and 5). In these experiments, the imaging time (20 seconds), and thus the contrast agent bolus duration, was the same at ASSET as at standard MR angiography, while the number of phase-encoding lines was doubled at ASSET to improve the spatial resolution. Within each of the two groups, the amounts and concentrations of the contrast agent were the same at ASSET as at standard MR angiography, because of the identical imaging times.


View this table:
[in this window]
[in a new window]

  		TABLE 2. ASSET MR Angiography with Acceleration Factor of Two to Increase Spatial Resolution

 
Data Analysis
Images were analyzed on a workstation with built-in software (Advantage version 4.0, GE Medical Systems). For all subjects, signal intensity was measured in four selected regions of interest as follows: right renal artery, aorta at the level of the renal arteries, soft tissue below the right renal artery, and background noise in the air outside the body (Fig 2). The size of the regions of interest was 50–70 mm2. Signal-to-noise ratio (SNR) was defined as SNR = SI/{sigma}N, where SI is the signal intensity of selected regions of interest and {sigma}N is the SD of background noise. Contrast-to-noise ratio (CNR) between the blood vessels (right renal artery and the aorta) and surrounding soft tissue was calculated as CNR = (SIV – SIT)/ {sigma}N, where SIV and SIT are the signal intensities measured in blood vessels and surrounding tissues, respectively, and {sigma}N is the SD of background noise. Two investigators (C.V.Q. and S.K.K.) who were blinded to the method of data acquisition analyzed the data independently, and the results were averaged to determine the final values. Maximum intensity projections (MIPs) were created directly from the contrast-enhanced images without the use of a mask data set.



View larger version (148K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Anterior reference image from contrast-enhanced MR angiography with a standard technique shows regions of interest selected for data analysis: right renal artery (1), aorta at the level of the renal arteries (2), soft tissue below the right renal artery (3), and background noise in the air outside the body (4). The image data set was acquired with a 3D spoiled gradient-recalled-echo sequence (oblique coronal plane, 5.1/1.3, 40° flip angle, 125-kHz bandwidth).

 
Statistical Analysis
Statistical analysis was performed by using a commercial software package (SPSS, version 11.0.1; SPSS, Chicago, Ill). For each of the five study groups, data obtained with ASSET and standard MR angiography were compared to determine whether there was a significant difference in SNR and CNR between these two imaging techniques. A paired-samples two-tailed t test was used, and a P value of .05 or less was considered to indicate a statistically significant difference. SNR and CNR were presented as means ± SDs.


   Results
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
Discussion
 REFERENCES
 
MIPs of the contrast-enhanced images obtained by using the standard MR angiography technique (imaging time, 20 seconds) and those acquired by using ASSET with an acceleration factor of two (imaging time, 10 seconds) are shown in Figure 3. Note that images obtained with ASSET show less venous enhancement because of the shorter imaging time and thus depict the left renal artery more clearly, compared with images obtained with standard techniques. The MIPs from standard MR angiography and from ASSET MR angiography with an acceleration factor of four (imaging time, 5 seconds) are shown in Figure 4. In Figures 3 and 4, comparable enhancement in the aorta and renal arteries was observed between the MIPs from standard and ASSET MR angiography at both acceleration factors. Figure 5 compares the MIP of the standard renal MR angiogram acquired in 20 seconds with matrix size of 256 x 128 and that of ASSET acquired in the same imaging time (20 seconds) but with higher spatial resolution in the phase-encoding direction (matrix size, 256 x 256). Overall, both ASSET and standard MR angiography clearly showed enhancement in the aorta and renal arteries, and there was very little degradation of image quality in the MIPs with ASSET.



View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a. Anterior MIP images from data sets obtained with a 3D spoiled gradient-recalled-echo sequence (oblique coronal plane, 5.1/1.3, 40° flip angle, 125-kHz bandwidth) and with (a) standard technique and 20-second imaging time or (b) ASSET and 10-second imaging time. Images depict abdominal vascular structures such as descending aorta (short arrow) and left and right renal arteries (long arrows) with comparable image quality. Note that the ASSET image shows less venous enhancement because of the shorter imaging time and thus depicts the left renal artery more clearly.

 


View larger version (157K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b. Anterior MIP images from data sets obtained with a 3D spoiled gradient-recalled-echo sequence (oblique coronal plane, 5.1/1.3, 40° flip angle, 125-kHz bandwidth) and with (a) standard technique and 20-second imaging time or (b) ASSET and 10-second imaging time. Images depict abdominal vascular structures such as descending aorta (short arrow) and left and right renal arteries (long arrows) with comparable image quality. Note that the ASSET image shows less venous enhancement because of the shorter imaging time and thus depicts the left renal artery more clearly.

 


View larger version (134K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a. Anterior MIP images from data sets obtained with a 3D spoiled gradient-recalled-echo sequence (oblique-coronal plane, 5.1/1.3, 40° flip angle, 125-kHz bandwidth) and with (a) standard technique and 20-second imaging time or (b) ASSET and 5-second imaging time. Images depict abdominal vascular structures such as descending aorta (short arrow) and left and right renal arteries (long arrows) with comparable image quality.

 


View larger version (121K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b. Anterior MIP images from data sets obtained with a 3D spoiled gradient-recalled-echo sequence (oblique-coronal plane, 5.1/1.3, 40° flip angle, 125-kHz bandwidth) and with (a) standard technique and 20-second imaging time or (b) ASSET and 5-second imaging time. Images depict abdominal vascular structures such as descending aorta (short arrow) and left and right renal arteries (long arrows) with comparable image quality.

 


View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5a. Anterior MIP images from data sets obtained with a 3D spoiled gradient-recalled-echo sequence (oblique-coronal plane, 5.1/1.3, 40° flip angle, 125-kHz bandwidth) and with (a) standard technique and a matrix size of 256 x 128 or (b) ASSET and higher spatial resolution (256 x 256) in the phase-encoding direction (left-to-right). Images depict abdominal vascular structures such as descending aorta (short arrow) and left and right renal arteries (long arrows).

 


View larger version (141K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5b. Anterior MIP images from data sets obtained with a 3D spoiled gradient-recalled-echo sequence (oblique-coronal plane, 5.1/1.3, 40° flip angle, 125-kHz bandwidth) and with (a) standard technique and a matrix size of 256 x 128 or (b) ASSET and higher spatial resolution (256 x 256) in the phase-encoding direction (left-to-right). Images depict abdominal vascular structures such as descending aorta (short arrow) and left and right renal arteries (long arrows).

 
Tables 3 and 4 summarize the SNR and CNR measured from the standard MR angiography data acquired in 20 seconds and from those obtained by using ASSET (groups 1–5). Data from ASSET in groups 1–3 were acquired in one half or one quarter of the standard imaging time while maintaining the same spatial resolution as at standard MR angiography, whereas data from ASSET in groups 4 to 5 were acquired in the same imaging time as at standard MR angiography, while the spatial resolution was doubled in the phase-encoding direction.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Comparison of SNR and CNR between Standard Technique and ASSET with Increased Imaging Speed

 

View this table:
[in this window]
[in a new window]

  		TABLE 4. Comparison of SNR and CNR between Standard Technique and ASSET with Twice the Spatial Resolution in the Phase-Encoding Direction

 
The SNRs and CNRs measured from the standard MR angiographic data acquired in 20 seconds and those from ASSET MR angiographic data obtained at the same matrix size (256 x 128) but at reduced imaging times (groups 1–3) are plotted in Figures 6 and 7. In general, slightly reduced SNR was found with data acquired by using ASSET, while CNR was about the same between standard and ASSET MR angiographic data. For example, Table 3 shows that the SNR and CNR in the right renal artery at ASSET MR angiography with fourfold acceleration in acquisition (imaging time, 5 seconds; group 3) were 43.4 ± 18.0 and 35.9 ± 16.3, respectively, which was very close to those obtained by using the standard 20-second MR angiographic acquisition (SNR, 46.5 ± 22.7; CNR, 36.6 ± 21.8). Overall, P values ranged from .034 to .529 for SNR and from .129 to .908 for CNR, which indicated that there were either small or no significant differences in SNR and no significant differences in CNR (Table 3).



View larger version (26K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6. Bar graphs show comparison of SNR between standard MR angiograms acquired in 20 seconds (light gray) and ASSET MR angiograms acquired with reduced imaging time (dark gray). Matrix size of 256 x 128 was used for all examinations. Top: SNR in the descending aorta. Bottom: SNR in the right main renal artery. Data are means ± SDs. Note that there is only slight reduction in SNR with ASSET compared with the standard technique.

 


View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7. Bar graphs show comparison of CNR between standard MR angiograms acquired in 20 seconds (light gray) and ASSET MR angiograms acquired with reduced imaging time (dark gray). Matrix size of 256 x 128 was used for all examinations. Top: CNR in the descending aorta. Bottom: CNR in the right main renal artery. Data are means ± SDs. Note that CNR is nearly equivalent on standard and ASSET angiograms.

 
The SNRs and CNRs measured from the standard MR angiographic data acquired in 20 seconds and the ASSET MR angiographic data acquired in the same imaging time but with twice the spatial resolution in the phase-encoding direction (256 x 256; groups 4 and 5) are shown in Figures 8 and 9. Relative to those obtained with standard MR angiography, data obtained with ASSET showed considerable reduction in both SNR and CNR. For example, the SNR and CNR of the aorta at ASSET MR angiography with twice the phase-encoding resolution (Table 4, group 5) were 15.7 ± 3.2 and 12.8 ± 3.2, respectively, which were much lower than those obtained by using standard MR angiography (SNR, 31.0 ± 9.4; CNR, 24.3 ± 8.4). Overall, P values for groups 4 and 5 were less than .05, which indicated a significant reduction in both SNR and CNR (Table 4).



View larger version (22K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 8. Bar graphs show comparison of SNR between standard MR angiograms acquired with a matrix size of 256 x 128 (light gray) and ASSET MR angiograms acquired with a matrix size of 256 x 256 (dark gray). Imaging time of 20 seconds was used for all examinations. Top: SNR in the descending aorta. Bottom: SNR in the right main renal artery. Data are means ± SDs. Substantial reduction in SNR is shown with ASSET.

 


View larger version (25K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 9. Bar graphs show comparison of CNR between standard MR angiograms acquired with a matrix size of 256 x 128 (light gray) and ASSET MR angiograms acquired with a matrix size of 256 x 256 (dark gray). Imaging time of 20 seconds was used for all examinations. Top: CNR in the descending aorta. Bottom: CNR in the right main renal artery. Data are means ± SDs. Substantial reduction in CNR is shown with ASSET.

 

   Discussion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion
REFERENCES
 
In theory, given the same experiment conditions, the SNR of an image acquired with ASSET MR angiography with reduced imaging time, in respect to that obtained with standard technique and full-Fourier acquisition, can be described by the following equation (15,24,27):

where R is the acceleration factor and g is the local geometric noise factor (g > 1.0). SNRASSET and SNRSTD represent the SNRs of images obtained with ASSET and standard techniques, respectively. The equation suggests a reduction of at least 30% and 50% in SNR for images acquired in one half and one quarter of the standard imaging time, respectively. However, since the duration of the contrast agent bolus at ASSET MR angiography with 10- and 5-second temporal resolution was one half and one quarter, respectively, of that at standard MR angiography, contrast agent concentration was two or four times that with the standard technique. With this approach, images obtained with ASSET produced SNR and CNR nearly equivalent to those obtained with standard MR angiography. This increase in contrast agent concentration for the accelerated ASSET MR angiography seemed to compensate for the losses in SNR and CNR typically associated with the use of parallel imaging techniques.

The use of ASSET to increase spatial resolution, on the other hand, should be dealt with more carefully. For example, when an acceleration factor of R is used to increase the spatial resolution in the phase-encoding direction while maintaining the same imaging time, the SNR of the acquired image in comparison with an image obtained with the standard technique with full-Fourier acquisition is described by the following equation (15,24):

where g is the local geometric noise factor (g > 1.0), and SNRASSET and SNRSTD represent the SNRs of images obtained with ASSET and standard techniques, respectively. In such a case, a reduction of at least 50% in SNR is expected with a doubling of of the spatial resolution in the phase-encoding direction. Unlike data from the experiments performed in groups 1–3, which were acquired with reduced imaging time and in which losses in SNR and CNR could be compensated by increased contrast agent concentration, data obtained with increased spatial resolution with ASSET may have significantly decreased SNR and CNR. For example, substantial reductions in both SNR and CNR were observed in groups 4 and 5. Such an approach may have disadvantages, if high SNR and CNR are required for diagnosis. However, reductions in SNR and CNR may not be important, because of the intrinsically high SNR and CNR of 3D contrast-enhanced MR angiographic images. Further study in patients with vascular disease is needed to determine if the use of ASSET to increase spatial resolution is beneficial.

We attempted to calculate g (using the equation in the preceding paragraph) for each subject in groups 4 and 5. In these groups, experimental conditions (imaging time, contrast agent dose, and injection rate) were kept identical for standard and ASSET MR angiography. The mean g in the aorta was 0.85 ± 0.16 for group 4 and 0.99 ± 0.21 for group 5. This was somewhat puzzling, since g theoretically should always be greater than 1.0. One possible explanation for this discrepancy between theory and our experiment results is that the ASSET and standard MR angiograms were obtained 30 minutes apart, and therefore subjects may have moved between the two imaging examinations. Such motion may have caused the location of the coils to be different relative to the subject’s body, and, as a result, affected local coil geometry. More rigorous investigation is warranted to fully understand this issue.

It is also important to note that our study was performed by using a rather small number of subjects (six per study group), which may explain the rather large SD found in both SNR and CNR. We believe that further study with a larger sample size is required to obtain more reliable statistical results.

In summary, our results show that a parallel imaging technique may be used to help improve both temporal and spatial resolution at contrast-enhanced 3D MR angiography. In particular, the use of ASSET to increase imaging speed is very promising because SNR and CNR with ASSET are nearly identical to those obtained with a standard 20-second technique. In addition, the reduction in imaging time also helped to reduce breath-holding time and potential contrast agent timing problems. This may be especially beneficial to patients with vascular disease, who tend to be older and in general may not be able to hold their breath for 20 seconds during abdominal or thoracic MR angiography. The application of ASSET to increase speed at peripheral vascular MR imaging, on the other hand, requires extra caution to avoid the risk of data acquisition before peak enhancement when the bolus-chase technique is used, especially for patients who have slow blood flow. One solution to the problem would be to perform imaging multiple times at the same location to make sure data are obtained when the bolus passes through the region of interest. These multiple data sets may even be averaged to help increase the SNR. The use of ASSET to increase spatial resolution was proved feasible as well, although SNR and CNR were reduced as a result. The loss in SNR and CNR with such an approach may not affect contrast-enhanced 3D MR angiography, because of the intrinsically high SNR and CNR of contrast-enhanced 3D MR angiograms. Further study in patients with vascular disease is needed to assess the applicability of these techniques for detection of pathologic conditions.


   ACKNOWLEDGMENTS
 
The authors thank Linda Pierchala, RN, and LaKeisha Dean for their assistance with subject recruitment.


   FOOTNOTES
 
Abbreviations: ASSET = array spatial sensitivity encoding technique, CNR = contrast-to-noise ratio, MIP = maximum intensity projection, SNR = signal-to-noise ratio, 3D = three-dimensional

Author contributions: Guarantor of integrity of entire study, Q.C.; study concepts, Q.C., R.R.E.; study design, Q.C., C.V.Q., W.L.; literature research, Q.C., C.V.Q.; clinical studies, C.V.Q., W.L., Q.C.; data acquisition, Q.C., W.L.; data analysis/interpretation, C.V.Q., S.K.K., Q.C.; statistical analysis, C.V.Q., S.K.K., Q.C.; manuscript preparation, C.V.Q., S.K.K., Q.C.; manuscript definition of intellectual content, Q.C.; manuscript editing, all authors; manuscript revision/review, Q.C., C.V.Q.; manuscript final version approval, Q.C.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion
 REFERENCES
 

  1. Prince MR, Narasimham DL, Stanley JC, et al. Gadolinium-enhanced magnetic resonance angiography of abdominal aortic aneurysms. J Vasc Surg 1995; 21:656-669.[CrossRef][Medline]
  2. Maki JH, Chenevert TL, Prince MR. Three-dimensional contrast-enhanced MR angiography. Top Magn Reson Imaging 1996; 8:322-344.[Medline]
  3. Prince MR, Chenevert TL, Foo TK, Londy FJ, Ward JS, Maki JH. Contrast-enhanced abdominal MR angiography: optimization of imaging delay time by automating the detection of contrast material arrival in the aorta. Radiology 1997; 203:109-114.[Abstract/Free Full Text]
  4. Krinsky G, Maya M, Rofsky N, et al. Gadolinium-enhanced 3D MRA of the aortic arch vessels in the detection of atherosclerotic cerebrovascular occlusive disease. J Comput Assist Tomogr 1998; 22:167-178.[CrossRef][Medline]
  5. Maki JH, Chenevert TL, Prince MR. Contrast-enhanced MR angiography. Abdom Imaging 1998; 23:469-484.[CrossRef][Medline]
  6. Lee VS, Rofsky NM, Krinsky GA, Stemerman DH, Weinreb JC. Single-dose breath-hold gadolinium-enhanced three-dimensional MR angiography of the renal arteries. Radiology 1999; 211:69-78.[Abstract/Free Full Text]
  7. Stillman AE. Contrast enhanced body magnetic resonance angiography. Int J Card Imaging 1999; 15:131-137.[CrossRef][Medline]
  8. Shetty AN, Bis KG, Kirsch M, Weintraub J, Laub G. Contrast-enhanced breath-hold three-dimensional magnetic resonance angiography in the evaluation of renal arteries: optimization of technique and pitfalls. J Magn Reson Imaging 2000; 12:912-923.[CrossRef][Medline]
  9. Ho VB, Foo TK, Czum JM, Marcos H, Choyke PL, Knopp MV. Contrast-enhanced magnetic resonance angiography: technical considerations for optimized clinical implementation. Top Magn Reson Imaging 2001; 12:283-299.[CrossRef][Medline]
  10. Lee VS, Martin DJ, Krinsky GA, Rofsky NM. Gadolinium-enhanced MR angiography: artifacts and pitfalls. AJR Am J Roentgenol 2000; 175:197-205.[Free Full Text]
  11. Roditi G. Contrast-enhanced magnetic resonance angiography. Br J Surg 2002; 89:817-820.[CrossRef][Medline]
  12. Sodickson DK, Manning WJ. Simultaneous acquisition of spatial harmonics (SMASH): fast imaging with radiofrequency coil arrays. Magn Reson Med 1997; 38:591-603.[Medline]
  13. Jakob PM, Griswold MA, Edelman RR, Sodickson DK. AUTO-SMASH: a self-calibrating technique for SMASH imaging—simultaneous acquisition of spatial harmonics. MAGMA 1998; 7:42-54.
  14. Griswold MA, Jakob PM, Chen Q, et al. Resolution enhancement in single-shot imaging using simultaneous acquisition of spatial harmonics (SMASH). Magn Reson Med 1999; 41:1236-1245.[CrossRef][Medline]
  15. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P. SENSE: sensitivity encoding for fast MRI. Magn Reson Med 1999; 42:952-962.[CrossRef][Medline]
  16. Bankson JA, Griswold MA, Wright SM, Sodickson DK. SMASH imaging with an eight element multiplexed RF coil array. MAGMA 2000; 10:93-104.
  17. Griswold MA, Jakob PM, Nittka M, Goldfarb JW, Haase A. Partially parallel imaging with localized sensitivities (PILS). Magn Reson Med 2000; 44:602-609.[CrossRef][Medline]
  18. Jakob PM, Griswold M, Hillenbrand C, Heidemann R, Hahn D, Haase A. High speed and high resolution cardiac MRI (parallel acquisition techniques and modular imaging). MAGMA 2000; 11:52-54.
  19. Bydder M, Larkman DJ, Hajnal JV. Generalized SMASH imaging. Magn Reson Med 2002; 47:160-170.[CrossRef][Medline]
  20. Wang Y. Description of parallel imaging in MRI using multiple coils. Magn Reson Med 2000; 44:495-499.[CrossRef][Medline]
  21. Golay X, Brown SJ, Itoh R, Melhem ER. Time-resolved contrast-enhanced carotid MR angiography using sensitivity encoding (SENSE). AJNR Am J Neuroradiol 2001; 22:1615-1619.[Abstract/Free Full Text]
  22. Bammer R, Keeling SL, Augustin M, et al. Improved diffusion-weighted single-shot echo-planar imaging (EPI) in stroke using sensitivity encoding (SENSE). Magn Reson Med 2001; 46:548-554.[CrossRef][Medline]
  23. Weiger M, Pruessmann KP, Boesiger P. Cardiac real-time imaging using SENSE—sensitivity encoding scheme. Magn Reson Med 2000; 43:177-184.[CrossRef][Medline]
  24. Weiger M, Pruessmann KP, Kassner A, et al. Contrast-enhanced 3D MRA using SENSE. J Magn Reson Imaging 2000; 12:671-677.[CrossRef][Medline]
  25. Sodickson DK, McKenzie CA, Li W, Wolff S, Manning WJ, Edelman RR. Contrast-enhanced 3D MR angiography with simultaneous acquisition of spatial harmonics: a pilot study. Radiology 2000; 217:284-289.[Abstract/Free Full Text]
  26. Weiger M, Pruessmann KP, Boesiger P. 2D SENSE for faster 3D MRI. MAGMA 2002; 14:10-19.
  27. Sodickson DK, Griswold MA, Jakob PM, Edelman RR, Manning WJ. Signal-to-noise ratio and signal-to-noise efficiency in SMASH imaging. Magn Reson Med 1999; 41:1009-1022.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 RadiologyHome page
C. Heilmaier, R. Sutter, A. M. Lutz, B. Seifert, D. Weishaupt, B. Marincek, and J. K. Willmann
Mapping of Hepatic Vascular Anatomy: Dynamic Contrast-enhanced Parallel MR Imaging Compared with 64 Detector Row CT
Radiology, December 1, 2007; 245(3): 872 - 880.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
R. Sutter, D. Nanz, A. M. Lutz, T. Pfammatter, B. Seifert, A. Struwe, C. Heilmaier, D. Weishaupt, B. Marincek, and J. K. Willmann
Assessment of Aortoiliac and Renal Arteries: MR Angiography with Parallel Acquisition versus Conventional MR Angiography and Digital Subtraction Angiography
Radiology, October 1, 2007; 245(1): 276 - 284.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
H. J. Michaely, H. Kramer, O. Dietrich, K. Nael, K.-P. Lodemann, M. F. Reiser, and S. O. Schoenberg
Intraindividual Comparison of High-Spatial-Resolution Abdominal MR Angiography at 1.5 T and 3.0 T: Initial Experience
Radiology, September 1, 2007; 244(3): 907 - 913.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Roentgenol.Home page
S. Potthast, G. M. Bongartz, R. Huegli, A.-C. Schulte, J. G. Schwarz, M. Aschwanden, and D. Bilecen
Intraarterial Contrast-Enhanced MR Aortography With and Without Parallel Acquisition Technique in Patients with Peripheral Arterial Occlusive Disease
Am. J. Roentgenol., March 1, 2007; 188(3): 823 - 829.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Neuroradiol.Home page
J.-Y. Gauvrit, X. Leclerc, C. Oppenheim, T. Munier, D. Trystram, H. Rachdi, F. Nataf, J.-P. Pruvo, and J.-F. Meder
Three-Dimensional Dynamic MR Digital Subtraction Angiography Using Sensitivity Encoding for the Evaluation of Intracranial Arteriovenous Malformations: A Preliminary Study
AJNR Am. J. Neuroradiol., June 1, 2005; 26(6): 1525 - 1531.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chen, Q.
Right arrow Articles by Edelman, R. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, Q.
Right arrow Articles by Edelman, R. R.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE