(Radiology. 2000;214:283-289.)
© RSNA, 2000
Vessel Tracking: Prospective Adjustment of Section-selective MR Angiographic Locations for Improved Coronary Artery Visualization over the Cardiac Cycle1
Thomas K. F. Foo, PhD,
Vincent B. Ho, MD and
Maureen N. Hood, BS
1 From the Applied Science Laboratory, GE Medical Systems, Room 110-MRI, Johns Hopkins Hospital, 600 N Wolfe St, Baltimore, MD 21287 (T.K.F.F.); the Department of Radiology, Uniformed Services University of the Health Sciences, Bethesda, Md (V.B.H., M.N.H.); and the Laboratory of Cardiac Energetics, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Md (V.B.H.). Received September 29, 1998; revision requested December 7; revision received April 12, 1999; accepted July 28. Address reprint requests to T.K.F.F. (e-mail: Thomas.Foo@med.ge.com).
 |
Abstract
|
|---|
To follow the motion of the coronary artery in magnetic resonance angiography, the authors evaluated vessel tracking, a method for prospective adjustment of the section location as a function of the delay from the cardiac trigger. In 10 volunteers and four patients, this method allowed the vessel to be maintained in the plane of acquisition throughout the cardiac cycle. With a single-phase multisection sequence, vessel-tracking acquisitions had an efficiency of 0.68 ± 0.04 for both the right and left coronary arteries compared with 0.19 ± 0.03 for a non–vessel-tracking acquisition (P < .001).
Index terms: Coronary vessels, MR, 54.12142 • Magnetic resonance (MR), motion correction, 54.12142 • Magnetic resonance (MR), vascular studies, 54.12142
|
Introduction
|
|---|
Existing methods for two-dimensional breath-hold magnetic resonance (MR) imaging of the coronary arteries include single-phase single-section (1–4), single-phase multisection (5), or multiphase single-section (6,7) sequences. These methods all rely on the assumptions that the coronary artery is at a known section position at the appropriate cardiac phase and that there is minimal movement or displacement of the vessel during diastole. From a manual prescription of the desired imaging plane, these approaches perhaps guarantee capturing of a segment of the coronary artery in at least one or two images. With use of these existing techniques, numerous repeat breath-hold acquisitions with variations in the section position or angle are usually necessary to visualize more complete segments of the coronary arteries. This is especially true if the vessel is tortuous.
The problems with having a patient undergo a coronary MR angiographic examination that requires repeated breath holds (at least 15–25) (8) have resulted in wider use of three-dimensional navigator-echo sequences that do not require the patient to hold his or her breath (9–13). However, success of such navigator-echo techniques to reduce the respiratory-motion artifacts depends on patient cooperation and a consistent respiratory rhythm (14). Breath holding, if well tolerated by the patient, will yield images with the fewest respiratory-motion artifacts (15).
Duerinckx and Atkinson (16) had earlier proposed a single-section, single-phase approach with the section position of each acquisition manually adjusted according to the phase of the cardiac cycle during which data were acquired in an attempt to follow the coronary artery motion. By playing the images from these multiple acquisitions, each at a different phase of the cardiac cycle, a cinelike view of the dynamic cardiac motion can be visualized, with the vessel of interest being approximately maintained in the imaging plane at all times. However, with this approach, only one image was acquired per breath hold.
Our proposed alternative method is to automate tracking of the coronary artery position as a function of the delay from the electrocardiographic R wave, for a single acquisition at every cardiac phase. In automation of the tracking of the vessel, the section-selective radio-frequency pulse-frequency offset is adjusted to prospectively follow the coronary artery as it moves during the cardiac cycle (vessel tracking). Rather than acquisition of a single image for each breath hold, multiple images are acquired in a single (breath-hold) acquisition, which allows tracking of the vessel position as a function of the cardiac phase. This increases the spatial coverage efficiency of a breath-hold image, measured as the ratio of the number of images in an acquisition in which a substantial length of the coronary artery is visualized to the total number of images generated in that acquisition. The proposed approach is a hybrid between a multiphase single-section cine acquisition (FASTCARD; GE Medical Systems, Milwaukee, Wis) in which the imaging location is fixed and a single-phase multisection (multiplanar) technique with multiple fixed imaging locations, each at a different cardiac phase. Our proposed method is for the acquisition sequence to follow the motion of the coronary artery and image that vessel at every point in the cardiac cycle, that is, to track the coronary artery motion. This allows for maximum spatial coverage of the coronary artery with a minimum number of breath holds.
|
Materials and Methods
|
|---|
Tracking of the motion of the coronary artery can be accomplished by using either a one-dimensional column excitation radio-frequency excitation pulse (or navigator echo) to track and correct the motion in real time, or by prospectively measuring the maximum excursion of the coronary artery with a multiphase cine scout acquisition in a plane orthogonal to the image acquisition plane, that is, a view in which the vessel can be visualized in cross section. In the latter case, the vessel position can be prospectively estimated from this view, allowing the imaging plane, with the vessel lying in plane, to track or follow the coronary artery as it moves from an end-diastolic position to an end-systolic position in systole and back again in diastole. By prospectively changing the section excitation position to track the motion of the vessel, spatial coverage of the coronary artery is increased over that with other more conventional two-dimensional acquisition techniques (Fig 1).
View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1. Diagram illustrates the maximum excursion of the right coronary artery as observed in a double-oblique plane tangential to the right atrioventricular (AV) groove with the right coronary artery viewed en face. The motion of the coronary artery is assumed to start from the end-diastolic position (bold solid line), move to the end-systolic position (dashed line) in systole, and return to the original position at the end of diastole. The maximum excursion defines the boundaries of the vessel motion during the cardiac cycle. The vessel-tracking algorithm then estimates the vessel position within these boundaries as a function of the cardiac cycle.
|
|
The maximum excursion of the vessel during the R-R interval, determined on the scout images, is identified as being from the end-diastolic position, Ddias, to the end-systolic position, Dsys (Fig 1). The position of the coronary artery is then constrained between these two locations. It is thus unnecessary to prescribe section locations beyond the boundaries determined by the maximum position excursions as observed on the cine scout image (Fig 2).
View larger version (201K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2. Multiphase single-section cine scout images show the proximal and distal right coronary artery (RCA) in cross section. On the images obtained during 15 phases of the cardiac cycle, the maximum displacement and the initial end-diastolic and end-systolic positions can be determined. The image acquisition plane would then be prescribed along a line that intersects the proximal and distal right coronary artery. The line on each image illustrates the extent of the spatial excursion of the right coronary artery throughout the cardiac cycle.
|
|
For the initial proof of concept, a linear velocity-motion model was used. Even though this only approximated the positional variation of the vessel motion, it permitted the section locations to straddle the true position, allowing better visualization of the different vessel segments. We used the model of Bacharach et al (17) to differentiate between motion during systole and that during diastole. The end-systolic time point (tsys) in milliseconds is then
where HR is the heart rate in beats per minute. From this expression, the time of diastole (tdias) can be easily calculated as tdias = R-Rtime - tsys.
If we assume a linear model, the velocities (v) are constant in systole and diastole such that
Note that higher order motion can also be included in the model. In addition, a persistence at end systole can also be incorporated into the model. However, these were not used in the proof-of-concept demonstration.
The section excitation position was then calculated on the basis of the delay time from the R wave. The displacement from the initial end-diastolic position D(i) is calculated thus,
otherwise, as a function of the segment number, i, where tacq(i) is the time where the ith k-space segment is acquired in the cardiac R-R interval. Note that for a segmented k-space acquisition, each segment of six to eight k-space views is assigned to a different section location (for single-phase multisection images) or a different cardiac phase (for multiphase single-section images).
Ten healthy volunteers and four patients (nine men and five women; age range, 31–67 years; mean age, 49 years) were enrolled in this project, which was approved by the institutional review board. Informed consent was received from all volunteers and patients. All studies were performed with a 1.5-T MR imager (Signa, version 5.6; GE Medical Systems, Milwaukee, Wis) outfitted with high-performance gradients (40 mT/m maximum amplitude and 150 T/m/sec maximum slew rate). All imaging sequences were variants of an electrocardiography-gated, fat-suppressed, two-dimensional, segmented k-space, gradient-echo pulse sequence with eight views per segment. In each subject, the right and left coronary arteries were localized by using a combination of sagittal, coronal, and transverse non–breath-hold single-phase multisection imaging and scout breath-hold standard multiphase single-section cine imaging performed in oblique planes perpendicular to the plane of maximal coronary artery motion (Fig 2). All breath holds were at end expiration to maintain better reproducibility in achieving comparable diaphragm locations in each acquisition (14).
On the breath-hold multiphase single-section cine images, the displacements of the targeted coronary artery could be visualized and measured. The breath-hold multiphase single-section cine image also served as a localizer image for prescribing the vessel-tracking, breath-hold single-phase multisection coronary angiographic sequence. The vessel-tracking acquisition parameters varied slightly according to subject size and heart rate and were in the range of repetition time msec/echo time msec of 6.9–9.8/1.9–2.5, 28–32-cm field of view, 5.0-mm-thick sections, 30°–40° flip angle, and 256 x 192 matrix. In seven subjects, breath-hold multiphase single-section cine coronary angiography was performed without vessel tracking in the same plane as was used in vessel-tracking imaging. At most, the localizer and scout acquisitions required no more than two to three breath holds. The initial sagittal and coronal images were acquired without breath holding in some of the studies, and this did not affect the ability to visualize the coronary artery segments.
On the scout breath-hold multiphase single-section cine images, the displacements of the coronary arteries were measured. Measured displacements were then compared with the calculated data from the vessel-tracking model. The coronary artery images with and those without vessel-tracking acquisition were evaluated independently by two observers (T.K.F.F., V.B.H.) in a blinded fashion to evaluate the efficiency of the acquisition. This efficiency was defined as the ratio of the number of images on which at least a 10-mm-long segment of coronary artery was visible to the total number of images per acquisition. All data were evaluated by using commercially available software (EXCEL; Microsoft, Redmond, Wash) and consisted of simple descriptive statistics and analysis of variance. A P value of less than .05 was considered statistically significant.
|
Results
|
|---|
The right coronary artery exhibited the greatest excursion. The distal right coronary artery had a mean displacement of 20 mm ± 4 (SD), whereas that for the proximal right coronary artery was 13 mm ± 5. The differences between the distal and proximal displacements indicate that the distal segments move at a higher velocity than do the proximal segments. In comparison, the proximal left main coronary artery had a mean displacement of 5 mm ± 1, as measured on a coronal scout acquisition, and the distal left main or left anterior descending coronary arteries had a mean displacement of 9 mm ± 4, as measured in the sagittal plane. The increase in the mean displacement as measured in the sagittal plane can be attributed to the inclusion of the distal portions of the left anterior descending coronary arteries, which have larger excursions.
Figure 3 shows the fit between the measured and estimated positions of the distal right coronary artery in two subjects, as a function of cardiac delay time. In all right coronary artery cases, the systolic excursions were better approximated than were the diastolic positions. Variations in the estimate of tsys from Equation (1) resulted in a shift between the estimated position of the coronary artery and that of the measured position.
View larger version (26K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3. Plots of estimated (*) and actual ( ) positional displacements of the distal right coronary artery as measured on a breath-hold cine scout image in two healthy volunteers, as a function of the cardiac delay time. In both cases, the vessel-tracking positions fit well in systole. In subject 1, however, the fit is less optimal in diastole. In subject 2, the estimated positions fit well in both systole and diastole.
|
|
Measurements for the left coronary artery (left main and left anterior descending arteries) were much better correlated with the linear model in both systole and diastole (as measured in the coronal or sagittal plane). As shown in Figure 4, the error between the predicted position of the vessel in both systole and diastole and the measured positions was less than a section thickness (5.0 mm). As the excursions for the left coronary artery tree are substantially less than those for the right coronary artery, the linear model is a better fit for this vessel segment.
View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4. Plots of estimated (*) and actual () positional displacements of the left main and proximal to middle left anterior descending arteries as measured in two different healthy volunteers on a breath-hold cine scout image in the coronal plane, as a function of the cardiac delay time. For subject A, the estimated positions fit well in both systole and diastole, within one-half the section thickness of the acquisition (±2.5 mm). For subject B, the motion fits well with a linear model, but the approximation of the time to systole is incorrect by almost 100 msec. The maximum displacement could also be adjusted for a better fit, which would lead to higher efficiency for vessel tracking.
|
|
As a measure of the efficiency of the acquisition, the ratio of the number of images on which a segment of the coronary artery was visualized to the total number of images in each acquisition was tallied for the two observers. For both the right and left coronary arteries combined, the efficiency (for both observers) for vessel-tracking acquisitions was 0.68 ± 0.04, whereas that for multiplanar, non–vessel-tracking acquisitions was 0.19 ± 0.03 (for both observers). The P value from an analysis of variance test between the two groups showed statistical significance (P < .001). Furthermore, in a comparison of the results for the two observers, the P value was .138, which indicates little interobserver variability.
The results were also analyzed for specific vessels. For the right coronary artery, the efficiency was 0.64 ± 0.05 with vessel tracking and 0.17 ± 0.01 with a standard single-phase multisection (interleaved multiplanar) non–vessel-tracking approach (P = .004). For the left main and left anterior descending coronary arteries, the efficiency was 0.73 ± 0.04 with vessel tracking compared with 0.20 ± 0.01 without vessel tracking in a multiplanar acquisition (P = .016). No statistically significant (P = .322) difference was noted between the efficiencies of the right and left coronary artery vessel-tracking acquisitions.
Figure 5 shows a vessel-tracking acquisition of the left anterior descending artery. The maximum excursion on these images was 20.0 mm. The left anterior descending artery was visualized on six of eight single-phase multisection vessel-tracking images (for an efficiency ratio of 0.75). In comparison, a similar single-phase multisection acquisition without vessel tracking (Fig 6) resulted in visualization of the left anterior descending artery on only two of eight images (efficiency ratio of 0.25). In addition, some of the images (three of eight) in the non–vessel-tracking acquisition were at section locations that were well beyond the position of maximum displacement of the vessel and hence had zero probability of depicting a vessel segment.
View larger version (169K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5. Vessel-tracking single-phase multisection images of the left anterior descending artery (LAD and arrow) in an oblique transverse plane in a 35-year-old man. The left anterior descending artery is depicted in six (frames 3 [top row, right] through 8 [bottom row, right]) of the eight images in this acquisition, with its approximately 104-mm length clearly delineated.
|
|
View larger version (172K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6. Conventional single-phase multisection images of the left anterior descending artery (LAD and arrow) with uniform 5-mm-thick contiguous sections in an oblique transverse plane in the same subject as in Figure 5. The left anterior descending artery is depicted in only two (frames 3 [top row, right] and 4 [middle row, left]) of the eight images in this acquisition. In addition, only an 83-mm length of the left anterior descending artery is well delineated. Note that three of the image locations (frames 6 [middle row, right] through 8 [bottom row, right]) are well beyond the position of the groove of the left anterior descending artery.
|
|
|
Discussion
|
|---|
Use of vessel tracking for coronary MR angiography can substantially improve visualization of different coronary artery segments and thereby minimize the number of breath-hold acquisitions required for sufficient spatial coverage. The vessel-tracking approach, in which the imaging plane is varied to follow the motion of the target vessel as a function of the cardiac cycle, demonstrated advantages over acquisitions at fixed spatial locations, such as the single-phase multisection or multiphase single-section cine techniques.
Alternatives to vessel tracking depend on the vessel of interest. The multiplanar approach of Meyer et al (5) is more suited for imaging of the right coronary artery, in which spatial excursion is large. For the left coronary artery, a multiphase single-section cine approach at a fixed single-section location is appropriate if the spatial excursions are less than 10 mm or twice the section thickness. In all cases, the proposed vessel-tracking technique offered improved spatial coverage for different coronary arteries over that with existing acquisition strategies, regardless of the spatial extent of the vessel excursion. This is because vessel tracking is better able to accommodate the full range of motion of the target vessel, unlike the single-phase multisection (multiplanar) or multiphase single-section cine approaches.
The increased spatial coverage also allowed a reduction in the number of breath holds required to image the coronary vascular tree. With the existing techniques, repeated adjustment of the imaging plane location is required for adequate visualization of contiguous sections of the coronary arteries. Recently, the vessel-tracking sequence has been further improved to obtain three separate overlapping vessel-tracking acquisitions in separate breath holds (Foo TKF, unpublished data, 1999). Based on the operator-defined vessel-tracking excursion, two additional acquisitions can be prescribed on either side of the original prescription, with the starting location for each acquisition offset 5 mm (one section thickness) above and below the original operator-defined end-diastolic position and with the same maximum excursion. This improvement was made in an effort to increase imaging efficiency and to accommodate for variations related to breath-hold position and to the nonlinear motion of the coronary arteries during diastole.
One additional benefit of vessel tracking is the greater amount of in-plane saturation provided in the immediate vicinity of the coronary artery because the number of radio-frequency pulses that the surrounding tissue will experience is greater. That is to say, although the tissue is moving, the spatially selective radio-frequency pulse is also moving synchronously. This could potentially improve vessel-to-background image contrast by decreasing background signal, a tool especially suited for imaging with a T1-shortening contrast agent.
The improved visualization with vessel tracking may also be realized when compared with multiphase single-section cine acquisitions. In Figure 7, the right coronary artery is depicted in 10 of 11 images obtained with vessel tracking (efficiency ratio of 0.91), in which the maximum displacement was 35 mm. In the multiphase single-section cine acquisition (Fig 8), the right coronary artery is seen in only the early systolic phases (ie, in two of 13 phases, or an efficiency ratio of 0.15). Owing to the motion of the vessel through the acquisition plane and the position of the section plane, only a 61-mm-long section of the right coronary artery was depicted with multiphase single-section cine imaging, as compared with 103 mm with the vessel-tracking single-phase multisection acquisition. Repositioning of the section location of the cine image allowed better visualization of the other vessel segments. However, this substantially increased the number of breath holds necessary. Typically, 15–25 breath holds are necessary for adequate spatial coverage of the coronary tree (8), not including the number of breath holds for the scout views. Furthermore, the temporal resolution of the multiphase single-section cine images may not be sufficient for viewing the vessel as it crosses a fixed imaging plane owing to differences in velocities of the proximal and distal coronary artery segments. Figure 9 shows a reformatted image from a three-dimensional navigator-echo acquisition with image quality comparable to that with the two-dimensional vessel-tracking acquisition but with a much longer imaging time (5 minutes).
View larger version (189K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7. Vessel-tracking images of the right coronary artery (RCA and arrow) in a 55-year-old man. In this example, 11 section locations were acquired in a single breath hold with segments of the right coronary artery depicted in 10 (frames 1 [top row, left] through 5 [middle row, left] and 7 [middle row, second from right] through 11 [bottom row, right]). Images are contrast material enhanced (MS-325 [0.05 mmol per kilogram of body weight]; Epix Medical, Cambridge, Mass). Acquisition parameters are 1.9/7.2, 32-cm field of view, 256 x 192 matrix, 30° flip angle, and one signal acquired. About 103 mm of the right coronary artery is depicted.
|
|
View larger version (190K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8. Single-section multiphase cine images of the right coronary artery (RCA and arrow) in the same patient as in Figure 7. Note that the segments of the right coronary artery are depicted in only two (frames 1 [top row, left] and 2 [top row, second from left]) of the 13 images. Furthermore, a smaller (61-mm) length of the right coronary artery is seen as the vessel moves through the imaging plane.
|
|
View larger version (157K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9. Prospectively gated three-dimensional navigator-echo reformatted image in approximately the same imaging plane and in the same patient as in Figures 7 and 8. Whereas the vessel-tracking two-dimensional acquisition was completed in about 1 minute, which allowed for a rest period between three 16-second breath holds, imaging time was about 5 minutes. Acquisition parameters are 4.9/1.8, 20° flip angle, 32-cm field of view, 24 2.0-mm-thick sections interpolated to 48 1.0-mm-thick sections, 256 x 192 matrix, and one-half signal acquired.
|
|
Although a linear model of motion was used in the experiments to demonstrate proof of concept, the variabilities of the motion in systole and diastole and from subject to subject make it difficult to implement a universal motion model applicable to all cases. On the basis of the data collected, it is clear that a linear model suffices for the motion during systole. However, there appears to be more variation in diastole with motion being more nonlinear and varying across individuals. This observation points to the need to tailor the motion in diastole on an individual basis. As an improvement in the concept, real-time adjustment of the motion model to attain an efficiency closer to 100% is a possible solution.
|
Acknowledgments
|
|---|
The authors thank Steven D. Wolff, PhD, and Robert S. Balaban, PhD, Laboratory of Cardiac Energetics, National Heart, Lung, and Blood Institute, for their support of this project.
|
Footnotes
|
|---|
Author contributions: Guarantors of integrity of entire study, T.K.F.F., V.B.H.; study concepts and design, T.K.F.F., V.B.H.; definition of intellectual content, T.K.F.F., V.B.H.; literature research, T.K.F.F.; clinical studies, T.K.F.F., V.B.H.; experimental studies, T.K.F.F., V.B.H.; data acquisition, T.K.F.F., V.B.H.; data analysis, M.N.H.; statistical analysis, T.K.F.F., V.B.H.; manuscript preparation, T.K.F.F.; manuscript editing, V.B.H., T.K.F.F.; manuscript review, T.K.F.F., V.B.H., M.N.H.
|
References
|
|---|
-
Edelman RR, Manning WJ, Burstein D, Paulin S. Coronary arteries: breath-hold MR angiography. Radiology 1991; 181:641-643.[Abstract/Free Full Text]
-
Manning WJ, Edelman RR, Li W. A preliminary report comparing magnetic resonance coronary angiography with conventional angiography. N Engl J Med 1993; 328:828-832.[Abstract/Free Full Text]
-
Manning WJ, Edelman RR, Boyle NG, Li W. Fat-suppressed breath-hold magnetic resonance coronary angiography. Circulation 1993; 87:94-104.[Abstract/Free Full Text]
-
Duerinckx AJ, Urman MK. Two-dimensional coronary MR angiography: analysis of initial clinical results. Radiology 1994; 193:731-738.[Abstract/Free Full Text]
-
Meyer CH, Hu BS, Nishimura DG, Macovski A. Fast spiral coronary artery imaging. Magn Reson Med 1992; 28:202-213.[Medline]
-
Sakuma H, Higgins CB, Foo TK, et al. Breath-hold MR cine angiography of coronary arteries in healthy volunteers: value of multi-angle oblique imaging planes. AJR 1994; 163:533-537.[Abstract/Free Full Text]
-
Hundley WG, Peshock RM, Hillis LD, et al. Noninvasive determination of infarct artery patency by cine magnetic resonance angiography. Circulation 1995; 91:1347-1353.[Abstract/Free Full Text]
-
Goldfarb JW, Edelman RR. Coronary arteries: breath-hold, gadolinium-enhanced, three-dimensional MR angiography. Radiology 1998; 206:830-834.[Abstract/Free Full Text]
-
Li D, Gutierrez FR, Kuginuki Y, et al. Coronary arteries: three-dimensional MR imaging with retrospective respiratory gating. Radiology 1996; 201:857-863.[Abstract/Free Full Text]
-
Post JC, Visser CA, Valk J, Hofman MB, van Rossum AC. Three-dimensional respiratory-gated MR angiography of coronary arteries: comparison with conventional coronary angiography. AJR 1996; 166:1399-1404.[Abstract/Free Full Text]
-
Achenbach S, Bachmann K, Laub G, et al. Visualization of the coronary arteries in three-dimensional reconstructions using respiratory gated magnetic resonance imaging. Coron Artery Dis 1997; 8:441-448.[Medline]
-
Kessler W, Bachmann K, Laub G, et al. Usefulness of respiratory gated magnetic resonance coronary angiography in assessing narrowings
0% in diameter in native coronary arteries and in aortocoronary bypass conduits. Am J Cardiol 1997; 80:989-993.[Medline]
-
Muller MF, Vock P, Meier B, Chatterjee T, Kroeker R, Fleisch M. Proximal coronary artery stenosis: three-dimensional MRI with fat saturation and navigator echo. JMRI 1997; 7:644-651.
-
Taylor AM, Pennell DJ, Firmin DN, Keegan J, Wiesmann F, Jhooti P. MR navigator-echo monitoring of temporal changes in diaphragm position: implications for MR coronary angiography. JMRI 1997; 7:629-636.
-
Hofman MBM, Paschal CB, Li D, Haacke EM, van Rossum AC, Sprenger M. MRI of coronary arteries: 2D breath-hold vs 3D respiratory-gated acquisition. J Comput Assist Tomogr 1995; 19:56-62.[Medline]
-
Duerinckx A, Atkinson DP. Coronary MR angiography during peak-systole: work-in-progress. JMRI 1997; 7:979-986.
-
Bacharach SL, Bonow RO, Green MV. Comparison of fixed and variable temporal resolution methods for creating gated cardiac blood-pool image sequences. J Nucl Med 1990; 31:38-42.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
S. B. Reeder, Y. P. Du, J. A. C. Lima, and D. A. Bluemke
Advanced Cardiac MR Imaging of Ischemic Heart Disease
RadioGraphics,
July 1, 2001;
21(4):
1047 - 1074.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Y. Wang, R. Watts, I. R. Mitchell, T. D. Nguyen, J. W. Bezanson, G. W. Bergman, and M. R. Prince
Coronary MR Angiography: Selection of Acquisition Window of Minimal Cardiac Motion with Electrocardiography-triggered Navigator Cardiac Motion Prescanning--Initial Results
Radiology,
February 1, 2001;
218(2):
580 - 585.
[Abstract]
[Full Text]
|
|
|
TOP
Abstract
Introduction
