| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Gasthuisberg University Hospital, Catholic University
of Leuven, Herestraat 49, B-3000 Leuven, Belgium.
2 ESAT-PSI, Gasthuisberg University Hospital, Catholic University of Leuven,
Herestraat 49, B-3000 Leuven, Belgium.
3 Philips Medical Systems, Veenpluis 4-6, 5680 DA Best, The Netherlands.
Received February 21, 2002;
accepted after revision April 1, 2002.
B. Giorgi is supported by a Marie-Curie Fellowship from the European
Commission.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Fifteen healthy volunteers underwent MR coronary angiography with a new balanced turbo field-echo sequence in comparison with the standard turbo field-echo sequence. Signal-to-noise, blood-to-myocardium, blood-to-fat, and blood-to-pericardial fluid contrast ratios of the left and right coronary artery systems were measured. Image quality was graded, the length and diameter of the coronary arteries were measured, and the number of visible side branches was assessed.
RESULTS. The balanced turbo field-echo images yielded a higher blood-to-myocardium and blood-to-pericardial fluid contrast ratio, a similar blood-to-fat contrast ratio, and a lower signal-to-noise ratio than the turbo field-echo images. On a 5-point grading scale (1, nondiagnostic or unreadable; 2, poor; 3, moderate; 4, good; 5, excellent), image quality was rated significantly better for the balanced turbo field-echo sequence than for the turbo field-echo sequence (left coronary artery, 4.0 ± 0.6 vs 3.6 ± 0.5 [p = 0.015]; right coronary artery, 4.4 ± 0.4 vs 3.6 ± 0.4 [p < 0.0001], respectively), resulting in a significantly longer segment of the three major coronary arteries visualized (left anterior descending coronary artery, 92 ± 21 mm vs 79 ± 24 mm; left circumflex coronary artery, 70 ± 7 mm vs 60 ± 18 mm; right coronary artery, 112 ± 28 mm vs 95 ± 27 mm) and a significantly higher number of side branches visualized (left anterior descending coronary artery, 2.9 ± 1.3 vs 1.5 ± 1.3; left circumflex coronary artery, 2.1 ± 1.7 vs 1.0 ± 1.2; right coronary artery, 3.7 ± 1.7 vs 2.6 ± 1.5). Mean imaging time per coronary artery was significantly shorter for the balanced turbo field-echo sequence (5.7 ± 1.0 min) than for the turbo field-echo sequence (8.4 ± 1.4 min) (p < 0.0001).
CONCLUSION. Compared with standard turbo field-echo MR coronary angiography, optimized balanced turbo field-echo MR coronary angiography improves the visualization of the coronary arteries and their side branches within a significantly shorter imaging time.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the recently published literature [9, 10], new technologic developments have been described comparing standard spoiled gradient-echo techniques with steady-state free-precession or balanced gradient techniques for acquisition of cine MR imaging. The term "balanced" refers to the refocusing of gradients in all three directions, causing a stable steady-state build-up that ensures maximum recovery of the transverse magnetization. The results of these studies showed a significant shortening of acquisition times and improved image quality, supplying images with less flow-dependent contrast between the ventricular lumen and the surrounding myocardium, as well as sharper delineation of the ventricular cavity.
The aim of our study was therefore to examine whether similar results could be achieved in imaging the coronary arteries by adapting the widely used 3D turbo field-echo MR coronary angiography technique to the balanced turbo field-echo technique. In our volunteer study, we compared signal-to-noise ratios, blood-to-myocardium contrast ratios, and blood-to-fat contrast ratios of the right and left coronary arteries. Image quality was graded, the length and diameter of the coronary arteries were measured, and the number of visible sides branches was assessed.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
MR Imaging
All images were obtained with a 1.5-T MR imaging unit (Intera CV; Philips
Medical Systems, Best, The Netherlands) equipped with master gradients
(amplitude, 30 mT/m; slew rate, 150 mT/m per millisecond) using a dedicated
cardiac software package on release 8 with the standard 5-element Synergy
cardiac coil (Philips Medical Systems) and VectorCardioGram (Philips Medical
Systems) trigering. All subjects were examined while in the supine position
with the VectorCardioGram leads on the anterior left hemithorax. The protocol
included six different images, all obtained during free breathing using
prospective respiratory navigator gating.
We used a standardized methodology to perform MR coronary angiography, which was previously described in the literature [6, 8, 11].
Scout Images
For all images, a local shim volume was positioned over the heart to
optimize the homogeneity of the magnetic field in this region. After
acquisition of the survey images in three orthogonal planes to localize the
heart within the chest, a low-resolution 3D balanced turbo field-echo
VectorCardioGram-triggered and respiratory navigator-gated scout image was
obtained in the transverse plane. The images acquired with this scan served to
identify large portions of the right and left coronary arteries for subsequent
positioning of the 3D volumes used for the submillimeter MR coronary
angiograms [8]. The parameters
of this sequence were as follows: TR/TE, 3.5/1.4; flip angle, 70°; field
of view, 290 mm; acquisition matrix, 159 x 208; and a rectangular field
of view, 100%. The k-space was filled by short balanced turbo field-echo shots
with a turbo factor of 25 resulting in an acquisition window of 91.5 msec. The
data were acquired during mid diastole by adapting the trigger delay to the
heart rate of the subject. The scan resulted in a 3D slab consisting of 40
slices with a reconstructed slice thickness of 2.5 mm and an in-plane
resolution of 1.39 x 1.82 mm. A T2-preparation pulse and a
fat-saturation pulse were applied immediately before data acquisition to
obtain high contrast between the blood and myocardium and between fat and
other structures, respectively
[12,
13]. Parallel imaging
technology (sensitivity encoding) was used to decrease the acquisition
duration of this image by a factor of 2
[14,
15].
MR Coronary Angiograms
The third through sixth angiograms were used for submillimeter 3D imaging
of the coronary arteries. The imaging slabs were positioned on the balanced
turbo field-echo scout images using the 3-point "planscan" tool, a
component of the extended cardiac software package, to coincide with the plane
of the coronary artery. Both turbo field-echo and balanced turbo field-echo
angiograms were obtained for the left and right coronary arteries. The imaging
parameters of both sequences are shown in
Table 1. The balanced turbo
field-echo images were acquired immediately after the turbo field-echo images,
using the previously described geometry and 3-point computed planning tool for
each artery. A schematic of the turbo field-echo and balanced turbo field-echo
imaging pulse sequences is shown in Figure
1.
|
|
Both sequences began with a flow-insensitive T2 prepulse for myocardial signal suppression followed by the navigator pulse and a spectrally selective fat-saturation pulse after which the turbo field-echo or the balanced turbo field-echo shot was acquired. The 3D slab consisted of 20 slices of 1.5 mm reconstructed thickness, resulting in coverage of 30 mm. Scans were reconstructed to a voxel size of 0.54 x 0.54 x 1.5 mm by additional zero-filling.
All 3D coronary scans used a radial phase-encoding profile order within the turbo shot with a "low-high" profile order. This method means that the first acquired echo in the turbo shot is in the central part of the k-space, and subsequent echoes are positioned on a more or less radial trajectory toward the periphery of the k-space [16]. Despite the pseudoradial filling, the spacing between all acquired profiles in the k-space along any straight line is equidistant. Also, an elliptic k-space shutter was applied (in the phase 2 encoding directions) to reduce scanning time by 20%, while maintaining good spatial resolution. Furthermore, "Constant Level AppeaRance" was used for all scans to correct for the nonuniform cardiac coil sensitivity (based on the separate reference scan that determines the coil sensitivity profiles for all cardiac coil elements and the same reference scan that is also used for sensitivity encoding in the scout scan).
MR Image Analysis
The MR coronary angiograms were transferred to a commercially available
workstation (Easy Vision; Philips Medical Systems) and analyzed. Different
signal intensities were measured on the native images to evaluate and compare
the effectiveness of the two techniques in distinguishing the coronary lumen
from the surrounding structures. The signal intensity of blood was measured
centrally in the lumen of the left main coronary artery and in the proximal
portion of the right coronary artery, approximately 1 cm beyond the origin,
with the region of interest confined within the boundaries of the vessel. To
obtain accurate measurements of the intracoronary signal, we magnified the
original image six times. The signal intensity of fat was measured in the
epicardial fat surrounding the proximal coronary arteries, and the signal
intensity for myocardium was measured in the left ventricular wall. The SD of
noise was calculated as the signal intensity of air divided by a factor of
1.25 [17]. The signal-to-noise
ratio was defined as the mean signal intensity from blood divided by the SD of
noise. The contrast ratio between blood and myocardium, blood and fat, and
blood and pericardial fluid was obtained by dividing their respective signal
intensities. The length and diameter of each coronary artery (right coronary
artery, left main coronary artery, left anterior descending coronary artery,
and left circumflex coronary artery) were evaluated. The diameter was measured
using calipers in the middle of the left main coronary artery and 1 cm beyond
the origin for the other coronary arteries. The length of the coronary
arteries was measured curvilinearly along its course.
For each coronary artery, image quality was visually graded using a 5-point scale, similar to scales used by other researchers to score the MR coronary angiograms: 1, nondiagnostic or unreadable; 2, poor; 3, moderate; 4, good; 5, excellent [8, 18]. Every vessel was divided into several segments of 3 cm each, and each segment was graded separately for image quality (e.g., 0-3 cm, 3-6 cm). The overall image quality for each coronary artery and the left and right coronary artery tree was obtained by averaging the results. The visual grading scale was based on assessment of vessel sharpness, motion artifacts, intracoronary vessel signal, and contrast between the coronary artery and the surrounding epicardial fat and myocardium. For instance, a grade 2 (poor) was assigned when a (curvi) linear structure located in the presumed position of the coronary artery was seen without clear depiction of vessel anatomy, whereas a grade 5 (excellent) was assigned when large portions of the coronary arteries were visible as sharply defined structures with visible branch vessels. The number of side branches for each coronary artery was counted for both sequences.
Statistical Analysis
The mean and SD of signal-to-noise ratios, blood-to-myocardium contrast
ratios, blood-to-fat contrast ratios, vessel length, diameter, and image
quality were calculated. The differences between the two protocols were
evaluated with a two-tailed paired t test for each coronary group of
data. A p value of less than 0.05 was considered to be statistically
significant. All results are shown as mean ± SD.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
|
Image Quality Assessment
Overall image quality of the left and right coronary artery trees yielded
significantly better image quality for the balanced turbo field-echo sequences
than for the turbo field-echo sequences: left coronary artery, 4.0 ±
0.6 (balanced turbo field-echo) versus 3.6 ± 0.5 (turbo field-echo)
(p = 0.015); and right coronary artery, 4.4 ± 0.4 (balanced
turbo field-echo) versus 3.6 ± 0.4 (turbo field-echo) (p <
0.0001) (Table 3). Although no
differences were found in image quality in the left main coronary artery
(p = 0.31) of the left coronary artery tree, there was a trend toward
better image quality for the balanced turbo field-echo technique in the left
anterior descending coronary artery (p = 0.059) and a significant
difference in the left circumflex coronary artery (p = 0.0024).
Segmental assessment of image quality for each coronary artery, as depicted in
Figure
3A,3B,3C,3D,
showed a different pattern between both sequences. For the turbo field-echo
technique, image quality gradually decreased in the distal direction, whereas
image quality for the balanced turbo field-echo was much less affected by the
distance along the coronary artery. Compared with the turbo field-echo
technique, the balanced turbo field-echo technique yielded significantly
better image quality in the mid coronary artery segments. In the distal
segments, the balanced turbo field-echo technique yielded significantly better
image quality in the right coronary artery. The small number of visualized
distal segments for the left anterior descending coronary artery and the left
circumflex coronary artery did not allow depiction of statistical differences
between the two techniques.
|
|
|
|
|
Visualization of the Coronary Artery Tree on MR Coronary
Angiography
In Figure
4A,4B,4C,
the diameter, length, and number of side branches for each coronary artery are
shown. The right and left main coronary arteries yielded slightly but
significantly smaller diameters for balanced turbo field-echo imaging than for
turbo field-echo imaging (Fig.
4A). For the coronary artery length, except for the left main
coronary artery, which was equally long on both sequences (balanced turbo
field-echo, 8.1 ± 3.2 mm vs turbo field-echo, 7.7 ± 3.1 mm;
p = 0.27), the other coronary arteries showed significant differences
in length. For the left anterior descending coronary artery, left circumflex
coronary artery, and right coronary artery, the balanced turbo field-echo
sequence showed longer segments than the turbo field-echo sequence: left
anterior descending coronary artery, 91.6 ± 21.3 mm versus 79.1
± 24.2 mm (p = 0.0053); left circumflex coronary artery, 70.5
± 17.6 mm versus 60.5 ± 17.9 mm (p = 0.0004); and right
coronary artery, 112.0 ± 27.7 mm versus 95.4 ± 27.0 mm
(p = 0.0033), respectively (Figs.
5A,5B,6A,6B,7A,7B,8A,8B,9A,9B).
The net percentage gain in length for the balanced turbo field-echo sequence
compared with the turbo field-echo sequence was 20.9% ± 28.5% for the
left anterior descending coronary artery, 14.6% ± 12.6% for the left
circumflex coronary artery, and 19.9% ± 24.7% for the right coronary
artery.
|
|
|
|
|
|
|
|
|
|
|
|
|
The number of coronary artery side branches detected by balanced turbo field-echo imaging was significantly higher than standard turbo field-echo imaging (Fig. 4C) in the left anterior descending coronary artery (p = 0.0006), left circumflex coronary artery (p = 0.0104), and right coronary artery (p = 0.0016). These side branches, when depicted by both techniques, resulted in a substantially longer extension and sharper definition on balanced turbo field-echo images (Figs. 6A,6B and 8A,8B)
MR Coronary Angiography Imaging Time
Balanced turbo field-echo imaging was also significantly faster (p
< 0.0001) than standard turbo field-echo imaging, the former achieving a
global mean scan time of 5.7 ± 1 min per coronary artery, whereas for
the latter, 8.4 ± 1.4 min per coronary artery was necessary. No
significant differences in scanning time (p = 0.35) were found
between the right and left coronary systems for the turbo field-echo sequence
or the balanced turbo field-echo sequence.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Earlier reports concerning cardiac cine MR imaging using the steady-state free-precession imaging approach and the breath-hold 3D true fast approach with steady-state precession MR coronary angiography have found a gain in signal-to-noise ratio compared with the turbo field-echo technique [9]. A comparison of the absolute signal-to-noise ratio values with the breath-hold MR coronary angiography techniques, however, shows higher signal-to-noise ratio values (left coronary artery, 12.0±3.7, and right coronary artery, 11.7±3.9, for the balanced turbo field-echo technique vs 9.0±2.7 for the breath-hold 3D true fast imaging with the steady-state precession approach [Shea S et al., presented at the Society for Cardiovascular Magnetic Resonance meeting, Orlando, FL, January 2002]).
In our study, we found a lower signal-to-noise ratio (a decrease of approximately 20%) for the balanced turbo field-echo sequence as compared with the regular turbo field-echo sequence. However, the difference in signal-to-noise ratio (and contrast-to-noise) between two different sequences is not absolute and can be controlled by many parameters that are chosen by the operator. In our study, the major difference between the balanced turbo field-echo technique and the regular turbo field-echo technique that directly impacted the signal-to-noise ratio was the readout bandwith, which, at a factor of 5.4, was larger for the balanced sequence (resulting in a lower signal-to-noise factor of 2.3, if nothing else had been changed). The impact of using this much larger bandwidth is partly compensated by the slightly larger acquisition voxel size and by the intrinsically higher signal-to-noise ratio of the balanced sequence, partly because it enables a larger flip angle. In retrospect, the signal-to-noise ratio of the balanced turbo field-echo sequence could have been increased using a somewhat smaller readout bandwidth with only a marginal penalty in TR and TE; in our study, a 20% increase in signal-to-noise could have been obtained (using a 500 Hz/pixel bandwith) with only an 8% increase in TR (5.4 instead of 5.0 msec). Furthermore, in our study, a simple zero-filling partial-echo reconstruction algorithm was used (for the regular turbo field-echo), which resulted in good signal-to-noise (at the cost of slight spatial blurring).
Equally important in MR coronary angiography is the contrast between intracoronary blood and surrounding tissues—that is, myocardium, epicardial fat, and pericardial fluid. Contrast between blood and myocardium shown on images obtained using turbo field-echo or gradient-recalled techniques is partly dependent on unsaturated spins of the blood inflowing in the slice of interest. This inflow enhancement is known to decrease with shorter TRs, which causes saturation of the blood signal and thus reduction of image contrast. The balanced-gradient technique is less dependent on inflow effects, and because TRs are shorter than the T2 values of the tissues in the imaging volume, the contrast is found to be inversely proportional to the ratio of T1 over T2.
In our study, this finding resulted in a superior blood-to-myocardium contrast ratio and consequently in a sharper depiction of vessel boundaries, mainly in the mid and distal segments of the coronary arteries and improved endocardial border definition in cine MR images as described by other authors [9]. Both imaging sequences use a spectral saturation pulse to suppress fat signal and to generate contrast between the coronary arteries and epicardial fat. A comparison of the blood-to-fat contrast ratios revealed no differences in the ratio for the left coronary artery system, and a slightly better contrast ratio for the balanced turbo field-echo sequence for the right coronary artery system was shown. The major benefit of an improved blood-to-fat contrast ratio is in the proximal part of the coronary arteries in which epicardial fat is abundantly present. More distally, however, coronary arteries are adjacent to the wall of the cardiac chambers and therefore are better depicted by the technique showing an improved blood-to-myocardium contrast ratio, especially for the mid and distal right coronary artery and most of the left circumflex coronary artery. Another advantage offered by balanced turbo fieldecho imaging, with its inherently high signal of fluids, was the clear differentiation of blood in the vessels from the pericardial fluid, which is more hyperintense than signal intensity of blood. In fact, with conventional turbo field-echo imaging, the distinction between blood and pericardial fluid was substantially difficult, particularly in association with those segments of coronary arteries that were surrounded by poor fat tissue [25].
The overall quality of images obtained for our study of balanced turbo field-echo imaging was good. Thus, balanced turbo field-echo imaging yielded a significantly higher score than turbo field-echo imaging. Qualitatively, visualization of the coronary artery (both the left and right system) was better with sharper vessel boundaries and a better visualization of the branch vessels. In the left coronary artery tree, the difference in image quality between the sequences was mainly due to a better image quality of the left circumflex coronary artery and a trend toward better image quality in the left anterior descending coronary artery, whereas image quality in the left main coronary artery was similar to both sequences. On turbo field-echo images, the image quality generally decreased the more distal the coronary artery segment. In contrast, the balanced turbo field-echo image often maintained a constant score throughout the segments.
A quantitative measurement of coronary artery diameters yielded slightly smaller values for the balanced turbo field-echo sequence than for the turbo field-echo sequence (statistically significant in the right coronary artery and the left main coronary artery). As discussed previously, balanced turbo field-echo imaging provides substantially sharper vessel boundaries with less blurring, which may explain the smaller diameters. A possible explanation for increased vessel sharpness might be the reduction in total scanning time, allowing a reduction in eventual motion artifacts, induced by the complexity and variation of the cardiac motion or respiratory patterns.
The most remarkable finding in this comparative study is the gain in visualization of coronary artery length using balanced turbo field-echo imaging. The gain is up to almost 21% compared with turbo field-echo imaging. Absolute measurements of the left coronary artery and right coronary artery yield superior values to most measurements reported in the literature [26, 27] and are close to the coronary are artery lengths reported by Li et al. [28]. As a result of the increased visualization of the coronary arteries in most volunteers, the right coronary artery was clearly visible in the inferior part of the right atrioventricular groove and in the proximal part of the inferior interventricular groove. The left circumflex coronary artery, known to be difficult to assess on MR coronary angiography, was clearly visible in all volunteers. These differences in the length of coronary artery between the two sequences can be related to the better blood-to-myocardium contrast ratio of balanced turbo field-echo imaging, yielding better vessel visualization in areas in which the coronary arteries are in close contact with surrounding soft-tissue structures other than epicardial fat.
Another interesting observation is the reduction in the total acquisition time (mean time < 6 min per MR coronary angiogram) using balanced turbo field-echo imaging. In our study, we noticed a gain of 32% for balanced turbo field-echo imaging compared with turbo field-echo imaging, without a loss in image quality or spatial resolution. This gain is obtained by the shorter TR of balanced turbo field-echo imaging that allowed using a higher turbo factor (16 instead of 10), with only a negligible increase in balanced turbo field-echo shot duration (87 msec instead of 71 msec). This improvement in acquisition times provides new opportunities for MR coronary angiography as a combined cardiac protocol in which other parameters such as ventricular function, myocardial perfusion, or late myocardial enhancement can be studied (Plein S et al., presented at the Society for Cardiovascular Magnetic Resonance meeting, January 2002).
Although balanced turbo field-echo imaging has several advantages, the sequence also has some limitations. As a result of a higher sharpness in the delineation of tissue interfaces (partly caused by susceptibility changes to which balanced turbo field-echo imaging is more sensitive than regular turbo field-echo imaging) and a higher signal intensity of noise, the overall appearance of images acquired with the balanced turbo field-echo technique is less smooth than that of images obtained with the corresponding turbo field-echo technique, with frequent exaltation of borders between different interfaces. One of the reasons why balanced turbo field-echo images appear relatively sharp is the fact that the signal falloff in the turboshot is less severe with balanced turbo field-echo than with the regular turbo field-echo. With the low—high profile order, as used in our study, this result means that for balanced turbo field-echo imaging, the peripheral k-space profiles (which by definition, determine the detail in the image) have a higher signal than those in standard turbo field-echo imaging. The signal intensity in the vessels measured on the balanced turbo field-echo images is not always as homogeneous as the one measured on the turbo field-echo images. This difference is particularly evident in the great vessels (aorta and pulmonary artery) and cardiac chambers, but the left main coronary artery might also experience this phenomenon. Although the balanced turbo field-echo sequence was constructed to be less sensitive to flow effects, spins moving in the blood vessels most likely create inhomogeneity that is not fully compensated by the balancing gradients in three directions and, hence, give balanced turbo field-echo imaging partial sensitivity to flow phenomena.
Further improvements in image quality, image resolution, signal-to-noise ratio, and contrast-to-noise ratios for MR coronary angiography can be achieved with the application of new acquisition schemes such as spiral imaging [29,30,31], which has a longer acquisition time compared with the radial k-space filling pattern or the use of contrast-enhanced MR coronary angiography. A major step to improve the quality and reliability of MR coronary angiography will be the advent of intravascular contrast agents in clinical practice. These blood pool agents, now used only in human trials, offer longer intravascular residence time allowing the increase of not only the vascular contrast, but also the optimal imaging window [32, 33].
In conclusion, we have shown that balanced turbo field-echo imaging provides an improved image quality throughout the coronary artery tree, with sharper contour delineation, superior visualization of the length of the coronary arteries, and depiction of side branches within a significantly shorter acquisition time.
The observed decrease in signal-to-noise ratio for the balanced turbo field-echo technique is not an intrinsic feature but is a result of controlled parameters like readout bandwidth, which in the case of the balanced turbo field-echo sequence, can be decreased to increase signal-to-noise ratio with only a marginal increase in TR and TE. The obvious advantages offered by the combination of a very short TR and the use of balanced gradients should allow balanced turbo field-echo imaging to be listed among the new pathways for MR coronary angiography and to be competitive with conventional turbo field-echo imaging that is currently in clinical use.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Ozgun, A. Hoffmeier, M. Kouwenhoven, R. M. Botnar, M. Stuber, H. H. Scheld, W. J. Manning, W. Heindel, and D. Maintz Comparison of 3D Segmented Gradient-Echo and Steady-State Free Precession Coronary MRI Sequences in Patients with Coronary Artery Disease Am. J. Roentgenol., July 1, 2005; 185(1): 103 - 109. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Jahnke, I. Paetsch, B. Schnackenburg, A. Bornstedt, R. Gebker, E. Fleck, and E. Nagel Coronary MR Angiography with Steady-State Free Precession: Individually Adapted Breath-hold Technique versus Free-breathing Technique Radiology, September 1, 2004; 232(3): 669 - 676. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Spuentrup, M. Katoh, A. Buecker, W. J. Manning, T. Schaeffter, T.-H. Nguyen, H. P. Kuhl, M. Stuber, R. M. Botnar, and R. W. Gunther Free-breathing 3D Steady-State Free Precession Coronary MR Angiography with Radial k-Space Sampling: Comparison with Cartesian k-Space Sampling and Cartesian Gradient-Echo Coronary MR Angiography--Pilot Study Radiology, May 1, 2004; 231(2): 581 - 586. [Abstract] [Full Text] [PDF]
|
|
|
|
