| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cardiac Imaging |
1 From the Departments of Radiology (T.S., M.H., U.H., A.S., W.A.W., S.F., F.T., H.S.) and Statistics and Biometrics (R.F.), University of Bonn, Sigmund-Freud-Strasse 25, D-53105 Bonn, Germany; Philips Medical Systems, Best, the Netherlands (J.G.); and the Department of Radiology, University of Pennsylvania, Philadelphia, Pa (H.L.). Received November 5, 2003; revision requested January 27, 2004; revision received June 3; accepted June 15. Address correspondence to T.S. (e-mail: t.sommer@uni-bonn.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: The study was approved by the institutional review board, and informed consent was obtained from all patients. Eighteen patients (11 men, seven women; mean age, 63 years; age range, 45–76 years) suspected of having coronary artery disease who were scheduled to undergo elective conventional coronary angiography (reference standard) were included. For coronary MR angiography at 3.0 and 1.5 T, a vector electrocardiographically gated three-dimensional segmented k-space gradient-echo imaging sequence was combined with real-time respiratory navigator gating and tracking. Signal-to-noise ratios (SNRs), contrast-to-noise ratios (CNRs), scores of image quality and sensitivity and specificity for the detection of coronary artery stenosis on a segment-by-segment basis were assessed at 3.0 and 1.5 T. Data were analyzed for statistical differences by using the Wilcoxon matched-pairs test and the McNemar test.
RESULTS: The average increase in SNR at 3.0 T with respect to that at 1.5 T was 29.5% for the left coronary artery (LCA) and 31.2% for the right coronary artery (RCA) (P < .001), and the average increase in CNR was 21.8% for the LCA and 23.5% for the RCA (P < .001). Scores of image quality (P = .77) and diagnostic accuracy for the detection of coronary artery stenoses (sensitivity and specificity: 82% and 89%, respectively, at 3.0 T vs 82% and 88% at 1.5 T; P > .99) were identical or almost identical at both field strengths.
CONCLUSION: Coronary MR angiography at 3.0 T is feasible in patients suspected of having coronary artery disease and yields significant increases in SNR and CNR, although current techniques do not result in significantly improved image quality and diagnostic accuracy compared with the quality and accuracy at 1.5 T.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Stuber et al (5) demonstrated the feasibility of coronary MR angiography at 3.0 T inhumans. However, this study was performed by using healthy volunteers, and the applicability of its results remains to be evaluated in the clinical context of patients suspected of having coronary artery disease. Furthermore, the efficacy for visualizing diseased coronary arteries and detecting coronary artery stenosis at 3.0-T imaging in direct comparison to the efficacy for these factors at 1.5-T imaging is as yet unknown.
Thus, the purposes of our study were to prospectively evaluate the feasibility, image quality, and accuracy of coronary MR angiography at 3.0 T in patients suspected of having coronary artery disease and to prospectively compare these results with those of coronary MR angiography performed at 1.5 T.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The mean weight of the patients was 75 kg (range, 51–98 kg). The mean heart rate (as assessed with electrocardiography at the beginning of the MR imaging examination) was 70 beats per minute ± 7.2 (standard deviation) at 3.0 T and 71 beats per minute ± 6.6 at 1.5 T (P = .76).
Exclusion criteria were the following: absence of a sinus cardiac rhythm, orthopnea, history of previous heart surgery, presence of coronary stent, and contraindications to MR imaging such as the presence of ferromagnetic implants. All patients underwent 3.0-T coronary MR angiography and 1.5-T coronary MR angiography. The examinations were performed in a random order within 24 hours of each other. All conventional coronary angiography examinations were performed within 21 days after the MR angiography examinations. This study was approved by our institutional review board, and informed consent was obtained from all 18 patients.
MR Imaging
The MR imaging examinations were performed with 3.0- and 1.5-T MR imaging units (Intera; Philips Medical Systems, Best, the Netherlands) that were equipped with identical gradient systems (maximal gradient amplitude, 30 mT; slew rate, 150 T/m/sec). The length of the magnet bore was 157 cm at 3.0 T and 160 cm at 1.5 T; the bore had an internal diameter of 60 cm at both field strengths. A body coil was used as the transmit coil at both field strengths (maximum field of view [FOV] at 3.0 T, 530 mm in left-right and anteroposterior directions and 400 mm in craniocaudal direction; maximum FOV at 1.5 T, 530 mm in left-right and anteroposterior directions and 480 mm in craniocaudal direction).
A six-element cardiac synergy surface receive coil with three flexible anterior and three rigid posterior rectangular (230 mm [left-right] x 320 mm [craniocaudal]) elements was used for imaging at 3.0 T, and a five-element cardiac synergy surface receive coil with two flexible anterior round (diameter, 200 mm) and three rigid posterior rectangular (138 mm [left-right] x 200 mm [craniocaudal]) elements was used for imaging at 1.5 T. Each coil was provided by the manufacturer (Philips Medical Systems) and represented the commercially available standard coil for cardiac imaging at that specific field strength. Coronary MR angiography data were analyzed at a workstation (EasyVision; Philips Medical Systems).
Three-dimensional MR Angiography
For coronary MR angiography, an electrocardiographically gated three-dimensional (3D) segmented k-space gradient-echo imaging sequence (turbo field echo) was combined with real-time respiratory navigator gating and tracking (6,7). Vector electrocardiographic R-wave triggering (8) was used at both field strengths and was combined with a heart rate–specific individual diastolic trigger delay to minimize intrinsic myocardial motion during the acquisition interval. All coronary data were acquired during free breathing by using the previously described "pencil-beam" navigator (9) localized at the lung-liver interface of the right hemidiaphragm with a 5-mm gating window and prospective adaptive 3D real-time motion correction of the imaged volume position. A constant correction factor of 0.6 was applied in the superior-inferior direction, and the navigator beam radius was 15 mm. The navigator acquisition at 3.0 T was modified to compensate for the increased susceptibility at the lung-liver interface according to the method of Stuber et al (5): The excitation duration of the navigator two-dimensional selective radiofrequency pulse was shortened by reducing the number of cycles in k-space from 12 at 1.5 T to eight at 3.0 T to minimize the sensitivity of the navigator to susceptibility and T2* effects.
The following two 3D MR angiography sequences were performed in all patients at both field strengths: a 3-cm-wide double oblique 3D sequence along the main axis of the right coronary artery (RCA) and a 3-cm transverse 3D sequence for visualization of the proximal and middle segments of the left coronary artery (LCA). Each 3D volume slab was segmented in 242 shots. Twenty sections with a section thickness of 3 mm and a 1.5-mm overlap were acquired. A radial 3D k-space acquisition scheme, with priority given to the lower ky and kz profiles, was used at both field strengths to accelerate data acquisition (10) by minimizing gradient-switching time. In this approach, the profiles of one R-R interval are acquired in a k-space plane (ky, kz), which is incrementally rotated around the kx axis with angular increments in each cardiac cycle (5).
Spatial resolution (true voxel size = 0.9 x 0.9 x 3.0 mm3, reconstructed to 0.7 x 0.7 x 1.5 mm3 through zero filling and interpolation), acquisition window (70 msec), echo time (1.8 msec), and acquisition time (3 minutes 29 seconds at a heart rate of 70 beats per minute, given a navigator efficiency of 100%) were identical for both field strengths. The repetition time at 3.0 T (5.7 msec) was slightly shorter than the repetition time at 1.5 T (6.0 msec). The radiofrequency excitation angles were 20° at 3.0 T and 30° at 1.5 T. This modification was necessitated by specific absorption rate (SAR) restrictions at 3.0 T to maintain a fixed short echo time and repetition time identical to those used in the 1.5-T protocol. This change also accounted for the prolonged T1 relaxation time of blood at 3.0 T (1550 msec [1] vs 1200 msec at 1.5 T [11]), which resulted in a decreased Ernst angle. The SAR value of the 3D MR angiography sequence was 0.9 W/kg at 1.5 T and 3.7 W/kg at 3.0 T. The receiver bandwidth was 172.4 Hz at 1.5 T and 197.4 Hz at 3.0 T.
A spectrally selective fat-saturation prepulse (time delay, 59 msec at both field strengths) was applied to enhance the contrast between the coronary blood pool and the surrounding epicardial fat, and a T2 preparation prepulse (echo time, 50 msec at both field strengths) was applied to enhance the contrast between blood and myocardium (12). The radiofrequency excitation angle of the fat saturation prepulse was lowered to 103° at 3.0 T (5) (versus 110° at 1.5 T) to account for the prolonged T1 of fat at 3.0 T (approximately 350 vs 270 msec at 1.5 T [as measured in-house]). Local volume shimming (100 x 100 x 100 mm) was performed at the level of the origin of the left and the right coronary arteries at both field strengths.
All patients (18 of 18) completed the coronary MR angiography examinations at both field strengths without complications.
Conventional Coronary Angiography
Conventional coronary angiography was performed at different institutions by experienced interventional cardiologists in multiple projections by using standard techniques (13). A quantitative angiographic analysis of the coronary arteries was performed at our institution by an experienced interventional cardiologist (K.T., with 5 years of experience) according to a standard algorithm (14). Digital cineangiograms were evaluated by using a calibrated off-line analysis package (Cardiovascular Angiography Analysis System mark II, or CAAS II; Pie Medical Imaging, Maastricht, the Netherlands). Three end-diastolic frames were analyzed to assess the severity of stenosis in two orthogonal planes. The percentage diameter stenosis of coronary artery stenoses was calculated with reference to the proximal normal vessel segment. A clinically significant stenosis was defined as a stenosis that was greater than 50% of the vessel diameter. The investigator was blinded to the MR angiography data.
Quantitative Image Analysis of Coronary MR Angiograms
Navigator acquisition efficiency, length of the visualized coronary arteries, and signal-to-noise and contrast-to-noise ratios were assessed in an identical manner at 3.0 T and 1.5 T by an investigator (M.H., with 3 years of experience in cardiac MR imaging) who was blinded to the field strength used.
Navigator acquisition efficiency.—The navigator acquisition efficiency of the coronary MR angiography sequence at 3.0 and 1.5 T was recorded in all patients separately for the LCA and the RCA. Navigator acquisition efficiency (given as a percentage) was defined as the number of shots accepted by the navigator divided by the total number of heartbeats it took to complete a scan.
Length measurements.—For length measurements of the coronary arteries, images of the left main (LM) coronary artery, the left anterior descending (LAD) artery, the left circumflex artery (LCX), and the RCA were reformatted in a previously described (15) multiplanar fashion by using the commercially available workstation that was previously mentioned.
SNR and CNR.—The SNR and CNR for each patient were calculated from the source images.
Regions of interest (mean size, 6.4 mm2 [13 pixels]; size range, 4.90–7.30 mm2 [10–15 pixels]) were placed in the LM coronary artery and in the proximal portion of the RCA. The standard deviation of the background signal intensity was measured in a larger region of interest (160–200 pixels) located anterior to the chest wall. The locations of the background regions of interest were chosen appropriately to avoid areas of phase ghosting.
SNR was determined as the mean value of signal intensity in the coronary arterial lumen divided by the standard deviation of the background signal intensity.
For measurements of the CNR between the coronary arteries and the myocardium, an additional region of interest (25 pixels) was located in the basal ventricular septum for the RCA and in the basal anterior wall of the left ventricle for the LCA.
CNR was defined as the mean value of the difference between the signal intensity in the coronary arterial lumen and the surrounding myocardium divided by the standard deviation of the background signal intensity.
All regions of interest were placed by the same author (M.H.) to ensure consistency.
Qualitative Image Analysis of Coronary MR Angiograms
Image quality.—The source images of the coronary MR angiography sequences at both field strengths were graded with respect to image quality independently by two radiologists (U.H., with 5 years of experience in cardiac MR imaging and A.S., with 3 years of experience). The investigators were blinded to the field strength used.
Artifacts were categorized as being attributable to normal MR image degradation related to ghosting or susceptibility effects.
After independent interpretations, discrepancies between the two readers in the scoring of vessel depiction were resolved by consensus to establish a final score. In cases in which no agreement could be reached, a final consensus diagnosis was obtained by using a third observer (T.S., with 10 years of experience in cardiac MR imaging) as an arbiter. In all cases, the readers were blinded to the field strength used.
Analysis of image quality was performed on a segment-by-segment basis by using the coronary MR angiography source images. The following seven coronary segments were evaluated for the presence of coronary artery stenoses: (a) the LM coronary artery, (b) the proximal (0–2.0-cm) segment of the LAD artery, (c) the middle (2.0–4.0-cm) segment of the LAD artery, (d) the proximal (0–1.5-cm) segment of the LCX artery, (e) the middle (1.5–3.0-cm) segment of the LCX artery, (f) the proximal (0–3.0-cm) segment of the RCA, and (g) the middle (3.0–6.0-cm) segment of the RCA. Image quality with respect to blurring of the coronary arteries and the presence of artifacts was graded by using a scale in which a grade of 1 indicated that the majority (
50%) of the coronary artery segment was not visible, was barely visible, or was obscured by major artifacts; a grade of 2, that a portion (<50%) of the coronary artery segment was not visible, was barely visible, or was obscured by major artifacts; a grade of 3, that the coronary artery segment was visualized without artifacts but had markedly blurred borders; a grade of 4, that the coronary artery segment was visualized without artifacts but had mildly blurred borders; and a grade of 5, that the coronary artery segment was visualized without artifacts and had sharply defined borders.
Coronary artery stenosis.—Two readers (M.H. and T.S.) reviewed the MR angiograms (the source images and the reformatted images) to assess the presence of hemodynamically relevant coronary artery stenoses. As at conventional coronary angiography, a clinically relevant stenosis was defined as stenosis of greaterthan 50% of the arterial diameter. The presence or absence of coronary artery stenoses greater than 50% at coronary MR angiography performed at 3.0 T and at 1.5 T was evaluated on a segment-by-segment basis. The seven coronary segments defined previously were evaluated for the presence of stenoses. The distal segments and side branches of the coronary arteries were not included in the analysis. Furthermore, coronary artery segments depicted on images with nondiagnostic image quality grades (grades 1 and 2) were excluded from stenosis evaluation. The readers were blinded to clinical data, to the results of conventional coronary angiography, and to the field strength used. Again, discrepancies between the two readers in assessing the presence of stenoses were resolved by consensus to establish a final consensus diagnosis. In cases in which no agreement could be reached, a final consensus diagnosis was obtained by using a third observer (U.H.) as an arbiter.
Statistical Analysis
Heart rate; navigator acquisition efficiency; SNRs, CNRs, and lengths of the visualized coronary arteries; and scores of image quality at 3.0 T and 1.5 T were compared and analyzed by using the two-tailed Wilcoxon matched-pairs test. To take into account that the image quality in the seven segments assessed in each patient might be correlated, means for the scores of image quality were first calculated for each patient. (These mean scores were used as input for the Wilcoxon matched-pairs test.) On the basis of these data, the mean and standard deviation of the image quality were calculated for all 18 patients and were compared on a patient-by-patient-basis between 3.0 and 1.5 T.
Sensitivity, specificity, and accuracy values for the detection of coronary artery stenosis were compared between field strengths by using the two-tailed McNemar test, with results of conventional angiography as the standard of reference. Ninety-five percent confidence intervals were calculated on the basis of results of logistic regression analysis performed by using generalized estimation equations to account for the correlation structure between several observations in one patient (ie, for clustering of stenoses in one patient). No statistical test was performed to analyze differences between male and female patients. Data were expressed as means ± standard deviations and—to take into account the use of nonparametric tests—as medians, 25% quantiles, 75% quantiles, and/or interquartile ranges (IQRs). The IQR was defined as the difference between the 75% quantile and the 25% quantile. P values of .05 or less were considered to indicate statistically significant differences.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Navigator efficiency.—Mean navigator efficiencies at 3.0 T and at 1.5 T, respectively, were 38% ± 5.0 (standard deviation) and 38% ± 5.6 for the LCA (mean difference, 0.1% ± 2.3; median difference, –1%; IQR, 4%; P = .82) and 37% ± 5.2 and 38% ± 5.0 for the RCA (mean difference, –0.1% ± 1.9; median, –1%; IQR, 3%; P = .64).
Length measurements.—The visualized coronary artery lengths at 3.0 T were not significantly different from those at 1.5 T for the LM artery (mean value, 17.2 mm ± 3.4 vs 16.8 mm ± 3.6; mean difference, 0.4 mm ± 0.9; median difference, 1.0 mm; IQR, 1.0 mm; P = .119) and the LCX artery (mean value, 46.9 mm ± 5.4 vs 46.8 mm ± 5.4; mean difference, 0.2 mm ± 1.3; median difference, 0 mm; IQR, 2.0 mm; P = .723). The visualized lengths were slightly higher at 3.0 T than at 1.5 T for the LAD artery (mean value, 62.2 mm ± 5.4 vs 61.5 mm ± 5.4; mean difference, 0.7 mm ± 1.3; median difference, 1.0 mm; IQR, 2.0 mm; P = .049) and the RCA (mean value, 103.6 mm ± 15.9 vs 102.8 mm ± 15.9; mean difference, 0.8 mm ± 1.3; median difference, 1.0 mm; IQR, 0 mm; P = .036).
SNR and CNR.—Mean SNRs and mean CNRs were significantly higher (P < .001) at 3.0 T than at 1.5 T. Mean SNRs at 3.0 T and 1.5 T, respectively, were 31.1 and 23.9 for the LCA and 28.7 and 21.8 for the RCA (Table 1). The average increase in SNR at 3.0 T with respect to the SNR at 1.5 T was 29.5% ± 10.3 for the LCA and 31.2% ± 9.8 for the RCA.
|
|
Mean image quality over all segments was equivalent at both field strengths, with scores of 3.48 ± 0.75 (95% confidence interval: 3.11, 3.86) at 3.0 T and 3.49 ± 0.76 (95% confidence interval: 3.12, 3.86) at 1.5 T (mean difference, –0.01 ± 0.2; median difference, 0; IQR, 0.14; P = .77). The scores for image quality are given in Table 3.
|
There were no statistically significant differences in the detection of coronary artery stenoses between coronary MR angiography at 3.0 T and coronary MR angiography at 1.5 T with respect to sensitivity (P > .99) and specificity (P > .99). The sensitivity and specificity values, respectively, for detecting coronary artery stenoses on a segment-by-segment basis with coronary MR angiography were 82% (14 of 17 segments) (95% confidence interval: 60%, 93%) and 89% (80 of 90 segments) (95% confidence interval: 78%, 94%) at 3.0 T and 82% (14 of 17 segments) (95% confidence interval: 58%, 94%) and 88% (80 of 91 segments) (95% confidence interval: 76%, 94%) at 1.5 T. Two of three false-negative and seven of 10 false-positive results were located in the same segments on both 3.0-T and 1.5-T images. The results are summarized in Table 4.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Potential impediments to high-field-strength cardiac imaging include increased radiofrequency power deposition (17), as SAR increases quadratrically with constant magnetic induction field (or B0), increased sensitivity to motion artifacts, and impaired R-wave triggering related to amplified magnetohydrodynamic effects (8). Furthermore, susceptibility-related local magnetic field variations (
B0) increase linearly with B0, which may result in difficulties in visualizing heavily calcified coronary arteries and/or in identifying the lung-liver interface for navigator gating.
To our knowledge, this is the first study in which the feasibility and accuracy of high-field-strength coronary MR angiography in patients suspected of having coronary artery disease were evaluated in direct comparison with data obtained in the same patients at 1.5 T. At both field strengths, we used a free-breathing navigator-based 3D k-space segmented coronary MR angiography technique with high spatial and temporal resolution, as well as special prepulses for suppression of fat and myocardial signal that have been well established in the evaluation of coronary artery disease at 1.5 T (6,15,16). This approach enabled the use of a short echo time, which is expected to reduce susceptibility distortions related to T2* effects in high-field-strength MR imaging. The sequence used for coronary MR angiography at 3.0 T in this study was adapted to the differing physical boundary conditions at a higher static magnetic field and was slightly modified compared with that used at 1.5 T.
One of the main limitations of the use of 3.0 T for MR imaging is radiofrequency-induced heating, as the SAR increases quadratrically with B0 (17). The coronary MR angiography approach evaluated in this study benefited from a short acquisition window during each R-R interval. However, keeping the other imaging parameters—including the short echo time/repetition time—identical to those used at 1.5 T necessitated lowering the excitation angle at 3.0 T to avoid SAR violation. Specifically, the radiofrequency excitation angle was lowered from 30° at 1.5 T to 20° at 3.0 T, both to limit radiofrequency power deposition and to account for the prolonged T1 relaxation time of blood at 3.0 T. The excitation angle of the fat saturation prepulse also was lowered, from 110° at 1.5 T to 103° at 3.0 T (5), to account for prolonged T1 relaxation time of fat at 3.0 T. Furthermore, the excitation duration of the navigator two-dimensional selective radiofrequency pulse was shortened to minimize the sensitivity of the navigator to susceptibility and T2* effects at the lung-liver interface (5).
Electrocardiographic gating and effective suppression of pulsation-induced motion in high-field-strength MR imaging is complicated by elevated T waves in the electrocardiogram related to amplified magnetohydrodynamic effects (8). Furthermore, increased susceptibility gradients at the lung-liver interface can interfere with navigator-based suppression of respiratory motion artifacts. Our data show that the scoring of image quality with respect to the blurring of coronary arteries and the presence or absence of artifacts was equivalent at 1.5 T and 3.0 T, indicating that both cardiac (vector electrocardiographic) and respiratory (navigator) motion suppression were as effective at 3.0 T as they were at 1.5 T. The feasibility of navigator gating at 3.0 T is further substantiated by the finding that navigator efficiency (defined as the number of shots accepted by the navigator divided by the total number of heartbeats used for completing a scan) at 3.0 T was equal to that at 1.5 T (P = .82 for the LCA and P = .64 for the RCA).
Another important finding in our study was a moderate but highly significant increase in SNR at 3.0 T compared with the SNR at 1.5 T (P < .001); this increase was 29.5% for the LCA and 31.2% for the RCA. Results of MR imaging–based studies of the brain at 3.0 T have shown that the theoretically predicted twofold gain in SNR at 3.0 T relative to that at 1.5 T can be achieved in vivo (18–20). The lower SNR gain at 3.0 T in coronary MR angiography in comparison with that at cranial MR imaging is probably related to more pronounced dielectric and conductive effects and the reduced penetration of radiofrequency excitation in the human thorax (21–23).
In the present study, the SNR gain in the coronary arteries at 3.0 T also resulted in a significantly higher CNR between blood and myocardium (21.8% increase in the LCA and 23.5% increase in the RCA) compared with the CNR at 1.5 T.
Our data compare favorably with SNR and CNR measurements at 3.0 T in the coronary arteries of healthy volunteers (5); SNR and CNR values, respectively, in these healthy volunteers were 34 and 19 for the left and 22 and 17 for the right coronary system. The reason for the slightly lower SNR and CNR values in our study is probably that our measurements were performed within the proximal coronary arteries themselves, whereas Stuber and colleagues (5) determined SNR and CNR in the aortic root at the level of the LM artery and the proximal RCA.
Our preliminary results demonstrate that the measured increases in SNR and CNR did not necessarily translate into improved accuracy for detection of stenoses. We hypothesize that the observed gain in SNR and CNR at 3.0 T may be too small to result in significantly improved depiction of coronary artery stenoses because other effects influencing image quality, including the efficiency of suppression of intrinsic and extrinsic motion artifacts, are dominant factors. Probably for the same reasons, the visualized coronary artery lengths at 3.0 T were not significantly different from those at 1.5 T for the LM and LCX arteries. Although the visualized lengths were significantly higher at 3.0 T for the LAD artery and the RCA, these differences were so small (within the millimeter range) that they can be assumed to be without any clinical relevance.
Further technologic advances leading to improved suppression of respiratory and cardiac motion artifacts (such as implementation of multiple navigators for individual correction of the imaged volume position in all three directions) and further adaptations of the imaging sequences to the physical conditions at 3.0 T (eg, optimization of the radiofrequency angle; optimization of the fat saturation prepulse, optimization of the T2 prepulse, and/or optimization of navigator excitations) may be necessary before coronary MR angiography can benefit fully from high-field-strength MR imaging techniques.
In summary, coronary MR angiography at 3.0 T is feasible in patients suspected of having coronary artery disease and yields significant increases in SNR and CNR, but current techniques do not result in significantly improved image quality and diagnostic accuracy compared with the image quality and accuracy achieved at 1.5 T.
The quality of images obtained with high-field-strength coronary MR angiography performed in patients examined in this preliminary study is promising, and further refinements in acquisition and postprocessing hardware and software can be expected to result in substantial advances for use of 3.0 T coronary MR angiography. The increase in SNR observed at 3.0 T may be beneficial in coronary MR angiography because it could be used in future investigations to improve spatial resolution or decrease imaging time—for example, by enabling the implementation of parallel imaging techniques.
One limitation of our study was the small number of patients with coronary artery stenoses examined at both field strengths. In addition, our study involved a specific coronary MR angiography approach, and other techniques, such as steady-state free precession imaging or contrast material–enhanced MR angiography, which would also benefit from the increase in T1 relaxation time, remain to be evaluated in high-field-strength coronary MR imaging.
Furthermore, it should be stated that a six-element surface-receive coil was used at 3.0 T rather than the five-element surface-receive coil used at 1.5 T to compensate for the reduction of radiofrequency penetration owing to the shorter wavelengths at higher field strengths and to achieve better radiofrequency sensitivity and homogeneity within the field of view. The specific effect of the slight differences in surface coil designs and dimensions at the different field strengths was not evaluated in this study.
In conclusion, the coronary MR angiography approach evaluated in this study—a free-breathing navigator-based 3D coronary MR angiography technique at 3.0 T—is feasible in a clinical setting—that is, in patients suspected of having coronary artery disease.
SNR and CNR at 3.0 T are improved significantly (P < .001) compared with those at 1.5 T, although the theoretically predicted twofold gain in SNR at 3.0 T was not achieved.
At the present early stage of technical development of high-field-strength coronary MR angiography, the observed improvements in SNR and CNR did not result in improved image quality or diagnostic accuracy for detection of coronary artery stenosis.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, T.S.; study concepts, T.S., M.H.; study design, T.S., M.H., U.H., J.G.; literature research, M.H., A.S., H.L.; clinical studies, M.H., A.S., U.H.; data acquisition, T.S., M.H., U.H., J.G.; data analysis/interpretation, T.S., M.H., U.H., W.W., S.F., F.T., A.S., H.L., H.S.; statistical analysis, R.F.; manuscript preparation, M.H., W.W., S.F., J.G.; manuscript definition of intellectual content, T.S., M.H., U.H.; manuscript editing, T.S., M.H., W.A.W.; manuscript revision/review, M.H., A.S., F.T., H.L., H.S.; manuscript final version approval, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. K. Kuhl, F. Traber, J. Gieseke, W. Drahanowsky, N. Morakkabati-Spitz, W. Willinek, M. von Falkenhausen, C. Manka, and H. H. Schild Whole-Body High-Field-Strength (3.0-T) MR Imaging in Clinical Practice * Part II. Technical Considerations and Clinical Applications Radiology, April 1, 2008; 247(1): 16 - 35. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Gharib, V. B. Ho, D. R. Rosing, D. A. Herzka, M. Stuber, A. E. Arai, and R. I. Pettigrew Coronary Artery Anomalies and Variants: Technical Feasibility of Assessment with Coronary MR Angiography at 3 T Radiology, April 1, 2008; 247(1): 220 - 227. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Dewey, F. Teige, D. Schnapauff, M. Laule, A. C. Borges, K.-D. Wernecke, T. Schink, G. Baumann, W. Rutsch, P. Rogalla, et al. Noninvasive detection of coronary artery stenoses with multislice computed tomography or magnetic resonance imaging. Ann Intern Med, September 19, 2006; 145(6): 407 - 415. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. K. Kuhl, P. Jost, N. Morakkabati, O. Zivanovic, H. H. Schild, and J. Gieseke Contrast-enhanced MR Imaging of the Breast at 3.0 and 1.5 T in the Same Patients: Initial Experience Radiology, June 1, 2006; 239(3): 666 - 676. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS |
