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


     

DOI: 10.1148/radiol.2311021237
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Movies
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kühl, H. P.
Right arrow Articles by Buecker, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kühl, H. P.
Right arrow Articles by Buecker, A.
(Radiology 2004;231:198-207.)
© RSNA, 2004

Cardiac Imaging

Assessment of Myocardial Function with Interactive Non–Breath-hold Real-time MR Imaging: Comparison with Echocardiography and Breath-hold Cine MR Imaging1

Harald P. Kühl, MD, Elmar Spuentrup, MD, Alexander Wall, MD, Andreas Franke, MD, Jörg Schröder, MD, Nicole Heussen, MSc, Peter Hanrath, MD, Rolf W. Günther, MD and Arno Buecker, MD

1 From Medical Clinic I (H.P.K., A.W., A.F., J.S., P.H.) and the Departments of Diagnostic Radiology (E.S., R.W.G., A.B.) and Medical Statistics (N.H.), University Hospital Aachen, Pauwelsstrasse 30, 52057 Aachen, Germany. Received September 29, 2002; revision requested December 10; final revision received September 20, 2003; accepted September 29. Address correspondence to H.P.K. (e-mail: hkuehl@ukaachen.de).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
PURPOSE: To compare a spiral gradient-echo sequence with a radial steady-state free precession sequence and to compare these two interactive real-time cardiac magnetic resonance (MR) imaging examinations with harmonic two-dimensional echocardiography (ECHO) for the evaluation of regional myocardial function.

MATERIALS AND METHODS: Electrocardiographically triggered breath-hold steady-state free precession (BH-SSFP) MR imaging was the reference standard. Thirty-five nonselected patients scheduled for routine ECHO were included. Data from corresponding two-, three-, and four-chamber long-axis views and a midventricular short-axis view were acquired with each modality. Image quality and depiction of segmental wall motion were scored semiquantitatively by using the 16-segment model of the American Society of Echocardiography. Repeated-measures analysis of variance was performed to assess differences in image quality and wall motion depiction scores among the four imaging methods. Agreement was assessed by using Cohen {kappa} statistics.

RESULTS: Compared with the image quality achieved with BH-SSFP MR imaging, the image quality achieved with radial MR imaging was similar (nonsignificant difference), but that achieved with spiral MR imaging and ECHO was significantly inferior (P < .0001). There were no significant differences in the image quality of the long- and short-axis views between the radial and BH-SSFP sequences, while the image quality of the long-axis spiral (P < .05) and the short- and long-axis ECHO (P < .0001) views was lower than that of the BH-SSFP views. Compared with the mean wall motion score for BH-SSFP MR imaging, the mean wall motion score for radial MR imaging was not significantly different, but those for ECHO (P < .05) and spiral MR imaging (P = .0003) were significantly lower. Cohen {kappa} coefficients for agreement with the BH-SSFP sequence regarding wall motion scoring were 0.47 for ECHO, 0.67 for the spiral sequence, and 0.89 for the radial sequence.

CONCLUSION: The radial sequence enables similar accurate assessment of regional wall motion compared with the BH-SSFP sequence and yields image quality that is superior to that yielded by the spiral sequence and ECHO.

Supplemental material: radiology.rsnajnls.org/cgi/content/full/2311021237/DC1

© RSNA, 2004

Index terms: Heart, MR, 511.121412, 511.121416, 511.12144, 511.12149 • Heart, US, 511.12981 • Magnetic resonance (MR), cine study, 511.12144 • Magnetic resonance (MR), comparative studies, 511.121412, 511.121416, 511.12144, 511.12149 • Magnetic resonance (MR), functional imaging, 511.121412, 511.121416, 511.12144, 511.12149


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Two-dimensional echocardiography (ECHO) is an established imaging examination for the assessment of segmental myocardial function. As a real-time imaging modality, it enables rapid image acquisition without electrocardiographic triggering and respiratorygating and yields excellent spatial and temporal resolution. Despite the introduction of technical developments such as harmonic ultrasonography (US) (1–3) and contrast material–enhanced ECHO (4,5), which has resulted in improved image quality, ECHO still has limited effectiveness in subjects with poor acoustic windows at imaging, such as patients with advanced chronic pulmonary disease, patients who have undergone open-chest cardiac surgery, and obese patients (6).

Cardiac magnetic resonance (MR) imaging is considered the reference standard for the evaluation of cardiac anatomy and function because it yields excellent image quality. One major drawback of using cardiac MR imaging, however, is the repeated breath holds and electrocardiographic triggering required for image acquisition. This limitation can be overcome by using real-time interactive cardiac MR imaging, which facilitates free-breathing, nontriggered image acquisition. The value of this imaging examination for the accurate assessment of myocardial volumes and global function has recently been demonstrated (7–11). Moreover, real-time interactive cardiac MR imaging has been shown to complement ECHO in the assessment of regional myocardial function (12). However, in that study, a gradient-echo MR imaging sequence was used. The effectiveness of the gradient-echo sequence is limited by the low contrast between blood and the myocardium and by flow artifacts, especially on the long-axis views of the left ventricle (13).

An interactive real-time steady-state free precession MR imaging sequence with radial k-space filling has become available (14,15). This sequence combines the advantages of the excellent image quality yielded by the steady-state free precession sequence and interactive real-time image acquisition. Another real-time sequence that is available with the MR imaging unit that we use is performed by using a gradient-echo sequence with spiral k-space filling.

The purpose of this study was to compare interactive real-time MR imaging performed by using a spiral gradient-echo sequence (ie, spiral MR imaging) with that performed by using a radial steady-state free precession sequence (ie, radial MR imaging) and to compare these two examinations with harmonic two-dimensional ECHO for the evaluation of regional myocardial function.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Study Patients
Thirty-five consecutive inpatients (14 women, 21 men; mean age, 62 years ± 13 [SD]; age range, 34–82 years) scheduled to undergo ECHO were included in this study. Exclusion criteria included previously performed pacemaker implantation and severe dyspnea (New York Heart Association class IV). Patients were not selected according to ECHO image quality. ECHO was performed for assessment of left ventricular function in 23 patients, for evaluation of valve disease severity in seven patients, for assessment of right ventricular function associated with severe pulmonary disease in two patients, for evaluation of suspected intracardiac masses in two patients, and for evaluation of constrictive pericarditis in one patient. Sixteen patients had normal left ventricular function, and 19 had reduced left ventricular function with regional wall motion abnormalities. Fifteen patients had coronary artery disease and had had a myocardial infarction. Twenty-seven patients had sinus rhythm, and eight had atrial fibrillation.

The mean body mass index of the study population was 28.1 kg/m2 ± 3.8 (range, 21.1–35.4 kg/m2). The age distributions of the male (mean age, 59 years ± 14) and female (mean age, 60 years ± 15) patients were not significantly different. In accordance with the requirements of the local ethics committee of our university hospital, each patient gave his or her informed consent to be examined before participating in the study.

Imaging Protocols
ECHO.—All ECHO examinations were performed by four experienced sonographers with long-standing experience (range, 5–11 years) performing ECHO. The examinations were performed with the patient in the left lateral decubitus position by using US machines (Sonos 5500, Philips Medical Systems, Andover, Mass; Sequoia, Siemens, Boulder, Colo) with harmonic imaging capability. Parasternal long- and short-axis views, as well as apical two-, three-, and four-chamber views, were acquired in each patient. Care was taken to optimize the transducer position for each ECHO view to achieve optimal image quality. In addition, gain and compression settings were optimized individually for optimal delineation of the endocardial borders. Three to five consecutive heartbeats were acquired from each view, which was stored on S-VHS videotape for data analysis.

MR imaging.—The mean time between the cardiac MR imaging and ECHO examinations was 1 day ± 2. Patients were examined while in the supine position with a 1.5-T whole-body MR imaging unit (Gyroscan ACS-NT; Philips Medial Systems, Andover, Mass), research software (INCA 2; Philips Medical Systems, Best, the Netherlands), and a gradient system with an amplitude of 21 mT/m and a slew rate of 100 mT/m/sec. A five-element phased-array cardiac coil was used for image acquisition. To facilitate the fastest image reconstruction, only the two anterior segments of the coil were used for data acquisition.

For interactive real-time cardiac MR imaging, two sequences were used: a gradient-echo sequence with spiral k-space filling (16) and a steady-state free-precession sequence with radial k-space filling (14,15). The imaging parameters used in the different sequences are summarized in Table 1. For real-time MR imaging, data were acquired during free breathing and without cardiac triggering. For spiral MR imaging, four spiral interleaves were acquired. Fat suppression was achieved by means of water-selective excitation. For radial undersampling, 80 radial k-space lines were acquired. The sliding window reconstruction technique (17) was applied for both imaging sequences. The real-time MR images were displayed online, at a frame rate of 10 images per second, at an operator console by using a dedicated real-time reconstructor (18). The positions of the image sections and the imaging parameters could be changed interactively on the fly (7).


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

  		TABLE 1. Technical Parameters Used in the Different MR Imaging Sequences

 
An electrocardiographically triggered segmented k-space cine BH-SSFP MR imaging sequence in which 20 heart-phase images per cardiac cycle were acquired was used as the reference-standard examination. Images in each of the four planes (ie, two-, three-, and four-chamber long-axis views and a short-axis view) were acquired during a single end-expiratory breath hold.

The standardized views acquired with ECHO were also acquired with the spiral, radial, and BH-SSFP MR imaging sequences. The short- and long-axis views were planned interactively in real time by using the spiral and radial sequences and with use of scout images obtained by using the BH-SSFP sequence. A vertical long-axis view was planned on an axial scout image. This imaging plane was used to plan a short-axis section perpendicular to the axis of the left ventricle. On these images, the imaging planes of the two-, three-, and four-chamber views were subsequently planned. The two-chamber view was oriented parallel to the interventricular septum to intersect the anterior and inferior walls. The three-chamber view was oriented to intersect the anterior septum and the posterolateral wall and included the left ventricular outflow tract, the aortic valve, and the mitral valve on one view. The four-chamber view intersected the inferior septum and the anterolateral wall, with careful exclusion of the left ventricular outflow tract.

On all of the long-axis views, the imaging plane was aligned to intersect the apex of the heart. In addition, a short-axis view oriented perpendicularly to the two- and four-chamber views was imaged at the level of the papillary muscle tips.

For radial and spiral MR imaging, a 5-second recording from each view was stored on videotape for analysis. For BH-SSFP MR imaging, digitized loops were stored on the hard drive of a computer workstation (EasyVision; Philips Medical Systems, Andover, Mass) for analysis. Real-time MR images were displayed online without the need for offline reconstruction. Interactive real-time planning and imaging of the different long- and short-axis views took, on average, less than 10 minutes per patient with the spiral and radial sequences and about 5 minutes with ECHO. BH-SSFP planning and imaging of the different planes took, on average, 12 minutes.

Data Analysis
The ECHO and cardiac MR imaging data were analyzed separately and independently by means of the consensus interpretation of two experienced observers, who had more than 2 years of experience evaluating regional myocardial function and were blinded to the patients’ clinical data. The ECHO data were analyzed by H.P.K. and J.S., and the cardiac MR imaging data were evaluated by E.S. and A.W. The MR images acquired by using the real-time sequences and the breath-hold sequence were analyzed at least 1 week apart. If consensus between the observers was not achieved, a third blinded and independent observer (A.F. for the ECHO data and A.B. for the MR imaging data) made the final decision regarding the segmental score. The 16-segment model of the American Society of Echocardiography (19) was used (Fig 1). For ECHO, the parasternal long- and short-axis views and the apical two- and four-chamber views were analyzed. In only the patients with very poor parasternal imaging windows that precluded data analysis, the apical three-chamber view was used to assess myocardial segments 1, 2, 9, and 10.



View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Drawing illustrates the segmentation model of the American Society of Echocardiography. CV = chamber view, SAX = short-axis view. Numbers 1-16 refer to the different myocardial segments. Note that segments 2, 4, 7, 10, 12, and 15 are represented twice—on the short- and long-axis views.

 
For cardiac MR imaging, the corresponding two- and four-chamber views and a midventricular short-axis section were evaluated. The three-chamber view was used to assess the segments that corresponded to the parasternal long-axis ECHO view.

Image quality was evaluated by using the three-point scale of Franke et al (20), with which image quality was judged according to how well the endocardium was visualized: A score of 2 meant good visibility of the endocardium; a score of 1, the endocardium was moderately or partly visible; and a score of 0, the endocardium was barely or not visible. For the cardiac MR imaging modalities (ie, with the spiral, radial, and BH-SSFP sequences), these scores were modified so that they also applied to the visibility of the myocardial wall. In addition, the depiction of segmental wall motion with each of the four imaging modalities was scored by using guidelines published by the American Society of Echocardiography (19): A score of 1 meant normal contraction; a score of 2, hypokinesia; a score of 3, akinesia; and a score of 4, dyskinesia. In cases of diverging scores for the segments evaluated twice (ie, on the long- and short-axis views), the higher score was considered. In cases of conflicting results regarding wall motion assessment, the view with the best endocardial visibility score was chosen.

Statistical Analyses
Data are expressed as means ± SDs. Because multiple scores were assigned for each patient, repeated-measures analysis of variance of the data obtained in 28 patients in whom all of the segments were depicted with all four imaging modalities was performed to assess global differences in visibility and wall motion depiction scores among the four imaging methods. A post hoc analysis also was performed to evaluate interactions between the different imaging modalities. By using the same procedures, differences in visibility and wall motion depiction scores between the different acquired long- and short-axis views were assessed separately, and differences in these scores among individual segments were assessed on a segment-by-segment basis. {alpha}-level Bonferroni adjustment for multiple pairwise comparisons was applied by using a global {alpha} error of 5% (ie, P < .05), which was considered to indicate statistical significance.

By using univariate analysis of regression, the relationships between endocardial visibility and wall motion score observed with each imaging modality were assessed separately. Cohen {kappa} coefficients were calculated as measures of global agreement between the three real-time imaging modalities (ie, spiral and radial MR imaging and ECHO) and the reference technique (BH-SSFP MR imaging) for wall motion analysis. {kappa} >= 0.7 was considered to indicate satisfactory agreement. A statistical software package (SPSS, version 10.0; SPSS, Chicago, Ill) was used to perform all statistical analyses.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Two patients refused to undergo cardiac MR imaging after entering the study. In addition, one patient was excluded from the final analysis because no radial MR imaging data were acquired. In four patients, selected views were not acquired with radial or spiral MR imaging (the three-chamber view in all four patients and the two-chamber view in two of the four patients) owing to problems recording on the videotape. Therefore, myocardial segments 1 and 9 on the three-chamber view and segments 6, 8, 14, and 16 on the two-chamber view, none of which is represented twice on the long- and short-axis views, had to be excluded from the analysis with these four patients. Thus, a total of 496 segments in 32 patients—16 segments in 28 patients, 14 segments in two patients, and 10 segments in two patients—were available for analysis with each of the four imaging techniques (Fig 2).



View larger version (24K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Flow diagram of patient recruitment. CMRI = cardiac MR imaging, pat. = patient.

 
Image Quality
Mean visibility scores are given in Table 2. Figure 3 shows the percentages of segments for which there was poor, moderate, or good image quality (in terms of segment visibility) with the different modalities investigated. With BH-SSFP MR imaging, there was good visibility of the majority (n = 465, 93.8%) of the 496 segments and moderate (n = 29, 5.8%) or poor (n = 2, 0.4%) visibility of a small proportion of the segments. At radial MR imaging, the percentages of segments associated with good (n = 470, 94.8%), moderate (n = 20, 4.0%), or poor (n = 6, 1.2%) visibility of the 496 segments were similar to those at BH-SSFP MR imaging (nonsignificant difference in image quality between BH-SSFP and radial MR imaging). With spiral MR imaging and ECHO, lower proportions of segments were associated with good image quality (262 [52.8%] and 280 [56.5%] segments, respectively) and considerable proportions of segments were associated with moderate (172 [34.7%] and 138 [27.8%] segments, respectively) or poor (62 [12.5%] and 78 [15.7%] segments, respectively) image quality (P < .0001 for difference in image quality between spiral MR imaging or ECHO and BH-SSFP MR imaging). The image quality results for ECHO and spiral MR imaging were not significantly different.


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

  		TABLE 2. Mean Visibility Scores for the Different Imaging Modalities

 


View larger version (16K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Bar graph illustrates the percentages of segments for which there was good (white bars), moderate (gray), or poor (black bars) image quality with the different imaging techniques. With the BH-SSFP and radial sequences, there was good depiction of the majority of segments, while image quality was impaired with the spiral sequence and ECHO.

 
Figure 4 shows the mean visibility scores for the different views acquired with the different imaging modalities. The mean visibility scores for the short-axis views acquired with the spiral, radial, and BH-SSFP sequences were not significantly different, whereas the mean score for ECHO was significantly lower than that for BH-SSFP MR imaging (P < .0001). There were no significant differences in the scores of the three long-axis views between the radial and BH-SSFP sequences. In contrast, for each of the three long-axis views, the image quality achieved with ECHO and the spiral sequence was inferior to that achieved with the BH-SSFP sequence. The visibility results for spiral MR imaging and ECHO were not significantly different.



View larger version (37K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Bar graph illustrates mean visibility scores for the short-axis (SAX) view and the long-axis two-, three-, and four-chamber views (CV) acquired by using ECHO (black bars) and radial (striped bars), spiral (white bars), and BH-SSFP (gray bars) sequences. The short-axis views had high visibility scores with all three MR imaging modalities but lower visibility scores with ECHO. The long-axis views had high visibility scores with the BH-SSFP and radial sequences but lower visibility scores with the spiral sequence and ECHO. * = P < .0001 for difference in scores between BH-SSFP MR imaging and either ECHO or spiral MR imaging. {dagger} = P < .05 for difference in scores between the BH-SSFP and spiral sequences.

 
Figure 5 shows the mean visibility scores for the 16 segments visualized with the four imaging modalities. Segmental analysis revealed no significant differences in visibility score between the radial and BH-SSFP sequences. Compared with BH-SSFP MR imaging, ECHO yielded significantly (see white bars in Fig 5, P < .0001) lower image quality in terms of depiction of the middle and apical lateral segments (ie, segments 12 and 13, respectively) and the anterior segments (ie, segments 14–16, respectively). Compared with the image quality achieved with BH-SSFP MR imaging (see white bars in Fig 5, P < .0001), the image quality achieved with spiral MR imaging was lower in terms of depiction of the basal anteroseptal and posterior segments (ie, segments 1 and 9, respectively), the apical septal and inferior segments (ie, segments 5 and 8, respectively), the basal and apical lateral segments (ie, segments 11 and 13, respectively), and the middle and apical anterior segments (ie, segments 14 and 16, respectively).



View larger version (43K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Bar graphs illustrate mean visibility scores for each of the 16 segments depicted with the different imaging modalities. Black bars indicate no significant differences in visibility scores between the given modality and BH-SSFP MR imaging. White bars indicate significant (P < .0001) differences in visibility scores between the given modality and BH-SSFP MR imaging.

 
Figure 6a and 6b are examples of diastolic and systolic stop-frame images of the heart that were obtained with the different imaging modalities in a patient with an old anterior myocardial infarction. The four movies (radiology.rsnajnls.org/cgi/content/full/2311021237/DC1) are cine movies of the three-chamber view of the same patient with the different imaging modalities. Figure 7 shows examples of heart images (long-axis two-chamber view and short-axis view) obtained in a patient with normal myocardial wall motion.



View larger version (164K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6a. (a) Diastolic and (b) systolic stop-frame images of the heart acquired with the four imaging techniques in a patient with an old anterior myocardial infarction. Shown from left to right are two-, three, and four-chamber long-axis views and the short-axis view acquired with BH-SSFP MR imaging (top row), radial MR imaging (second row), spiral MR imaging (third row), and ECHO (bottom row). Note the poor image quality of the long-axis spiral views.

 


View larger version (167K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6b. (a) Diastolic and (b) systolic stop-frame images of the heart acquired with the four imaging techniques in a patient with an old anterior myocardial infarction. Shown from left to right are two-, three, and four-chamber long-axis views and the short-axis view acquired with BH-SSFP MR imaging (top row), radial MR imaging (second row), spiral MR imaging (third row), and ECHO (bottom row). Note the poor image quality of the long-axis spiral views.

 


View larger version (154K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7. Corresponding diastolic (images in left column on left and right sides) and systolic (images in right column on left and right sides) stop-frame images of long-axis two-chamber views (left group of images) and short-axis views (right group of images) obtained in a patient with normal myocardial wall motion. BH-SSFP (top row), radial (second row), spiral (third row), and ECHO (bottom row) images are shown. Note the poor image quality of the long- and short-axis ECHO views.

 
Myocardial Wall Motion Analysis
The mean myocardial wall motion depiction scores for the different imaging modalities are listed in Table 3. There was no significant difference in mean wall motion score between the radial and BH-SSFP sequences, whereas ECHO (P < .05) and spiral MR imaging (P = .0003) yielded significantly lower mean scores than BH-SSFP MR imaging. The wall motion scores for ECHO and spiral MR imaging were not significantly different. When the wall motion scores were analyzed according to the different acquired views, no significant differences in score between the short-axis and two-chamber long-axis views were observed. Compared with the BH-SSFP views, the three-chamber ECHO (P < .05) and spiral (P = .0006) views had significantly lower mean wall motion scores, while only the four-chamber view acquired with the spiral sequence had a significantly lower mean score (P = .0014). Segmental analysis revealed no significant differences in wall motion scores among the different imaging modalities.


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

  		TABLE 3. Mean Wall Motion Scores for the Different Imaging Modalities

 
Univariate regression analysis revealed a weak but significant relationship between visibility score and wall motion score with ECHO (r2 = 0.21, P < .0001) and spiral MR imaging (r2 = 0.13, P < .0001) but not with radial (r2 = 0.03, nonsignificant difference) or BH-SSFP (r2 = 0.001, nonsignificant difference) MR imaging. Agreement regarding wall motion depiction was excellent ({kappa} = 0.89) between BH-SSFP MR imaging—the reference-standard examination—and radial MR imaging, but lower between the BH-SSFP and spiral examinations ({kappa} = 0.67) and between the BH-SSFP and ECHO examinations ({kappa} = 0.47). However, when only the segments that were associated with good image quality were compared between each of the real-time imaging modalities and the reference technique, high {kappa} coefficients were achieved at each comparison, indicating good agreement ({kappa} = 0.75 with ECHO, {kappa} = 0.85 with spiral sequence, and {kappa} = 0.92 with radial sequence). The magnitudes of disagreement between the real-time imaging modalities and the reference technique regarding segmental wall motion scores are shown in Table 4.


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

  		TABLE 4. Differences in Segmental Wall Motion Scoring between BH-SSFP MR Imaging and the Real-Time MR Imaging Modalities

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
REFERENCES
 
Two-dimensional ECHO is an established real-time imaging modality for the evaluation of regional myocardial function. The introduction of technical developments such as harmonic US (in which US signals are transmitted at a particular frequency and received at twice that frequency) and contrast-enhanced ECHO has led to improved two-dimensional image quality and increased sensitivity in the detection of wall motion abnormalities, as compared with the image quality and sensitivity achieved with fundamental techniques in which signals are transmitted and received at the same frequency (2,4,19). It is estimated, however, that a complete evaluation of segmental myocardial function cannot be performed in up to 10% of patients owing to limitations of the ECHO imaging window. This is a particular concern in patients with advanced pulmonary disease, with high body mass index, and who have undergone thoracic surgery (6). Real-time interactive MR imaging of the heart may be a valuable alternative imaging modality for these patients. With this modality, the excellent image quality of MR technology is combined with real-time image acquisition, obviating repeated breath holding or electrocardiographic gating.

In the present study, we achieved excellent image quality with real-time interactive cardiac MR imaging performed by using a radial sequence, which yielded endocardial visibility and wall motion results that were similar to those of BH-SSFP MR imaging, the reference technique, and superior to those of spiral MR imaging and ECHO.

The better performance of the radial sequence, as compared with the spiral sequence, in terms of wall segment depiction may be due to different reasons. With the gradient-echo sequence used with spiral imaging, the contrast between the blood pool and the myocardium is dependent primarily on blood flow, which results in reduced contrast in areas with low velocity caused by saturation (12,21) and in patients with reduced left ventricular function (22). This reduced contrast is especially problematic on long-axis views of the heart, on which flow is directed along the imaging plane. In contrast, on short-axis views, flow is directed perpendicularly to the imaging plane, resulting in constant refreshment of the spins and improved contrast between the myocardium and the blood. This improved contrast is demonstrated by the good image quality of short-axis spiral views. Furthermore, blurring along the endocardial border due to longer repetition and echo times and the pronounced sensitivity to offline resonance and susceptibility artifacts may be the reasons for the reduced image quality of spiral MR images (23,24).

With the steady-state free precession sequence used in radial and BH-SSFP MR imaging, the contrast between the blood pool and the myocardium is dependent on T2 and T1 properties, which remain constant during the cardiac cycle. Therefore, radial MR imaging is less sensitive to the effects of blood flow compared with gradient-echo MR imaging (21). This lower sensitivity results in improved image quality, especially on long-axis views (13), and explains the excellent agreement between radial and BH-SSFP imaging results.

Another reason for the better performance of radial MR imaging compared with spiral MR imaging is the fact that radial streak artifacts associated with undersampling in radial imaging are oriented perpendicularly to the endocardial-epicardial border, which enhances the delineation of the border (14,15). In this study, the delineation of the border was less well defined with spiral imaging. The nominal spatial resolution achieved with the two real-time MR imaging sequences, which is higher as previously reported for other real-time imaging sequences (9,25,26), was comparable to the spatial resolution achieved with the BH-SSFP sequence.

The low temporal resolution of both real-time MR imaging sequences is the main limitation that probably precludes the reliable assessment of segmental wall motion in patients who have heart rates of greater than 100 beats per minute and/or are experiencing physical or pharmacologic stress. Lowering the number of radial k-space lines for image reconstruction might be an effective way to increase the temporal resolution without sacrificing spatial resolution at radial imaging (15,27).

Despite the use of harmonic US, the image quality achieved with ECHO was judged to be poor for 15.7% of the segments. Although this may be a higher proportion of segments than expected, this number might reflect our patient population, which included obese patients and two patients with severe pulmonary disease, in whom the ECHO results were virtually nondiagnostic. We did not administer left-heart contrast agents, the use of which might have resulted in improved delineation of the endocardial border (4) and fewer poorly defined segments. However, when only those segments with good visibility scores were considered, agreement for wall motion assessment with ECHO was good.

Thus, because of its low cost, ubiquitous availability, ease of use, and rapidity, two-dimensional ECHO is considered the first-line imaging technique of choice for the evaluation of segmental myocardial function in routine clinical practice. In the future, real-time cardiac MR imaging may be a valuable alternative to ECHO in patients in whom the assessment of segmental myocardial function (at rest or during stress) would contribute to treatment but cannot be performed with ECHO because of a poor imaging window.

In a few studies (12,25,26), real-time cardiac MR imaging has been compared with ECHO or conventional breath-hold MR imaging techniques in the evaluation of regional myocardial function. To our knowledge, this is the first study in which investigators directly compared the currently available state-of-the-art technologies for two-dimensional ECHO with conventional and real-time cardiac MR imaging modalities for the assessment of regional myocardial function. The results confirm and further extend the findings of Yang et al (12), who compared two-dimensional ECHO with real-time spiral gradient-echo cardiac MR imaging for the evaluation of segmental wall motion in both patients with good and those with poor image quality. The authors observed similar agreement ({kappa} = 0.79) regarding segmental wall motion depiction between ECHO and real-time spiral cardiac MR imaging in the patients with good ECHO image quality. In the patients with poor ECHO image quality, real-time MR imaging was superior for the visualization of wall segments.

However, the Yang et al (12) investigation differs from the present study in two important aspects: First, in their study, ECHO examinations were performed in the conventional imaging mode, which is known to yield inferior image quality in difficult-to-image patients compared with the image quality achieved by using the harmonic US mode. Thus, the ECHO results may have been underestimated relative to the cardiac MR imaging results in these patients. Second, the authors did not use a reference-standard examination for wall motion assessment, so there is uncertainty about the true wall motion status of the different segments, particularly in view of the low temporal resolution of real-time MR imaging sequences.

In another study, Plein et al (11) compared real-time gradient-echo cardiac MR imaging with conventional breath-hold gradient-echo cardiac MR imaging for the assessment of segmental myocardial function. These authors observed excellent agreement between the two sequences, as demonstrated by a {kappa} coefficient of 0.896. In contrast to the MR imaging sequences used in the present study, Cartesian k-space filling combined with an echo-planar imaging readout was used for real-time MR imaging in the Plein et al study. This sequence is prone to chemical shift and susceptibility artifacts, which may contribute to regional degradation of image quality. The radial sequence is less sensitive to T2* artifacts because of the short repetition and echo times. Moreover, in the Plein et al study (11), the reconstruction and display of the images were not performed in real time and were delayed until after image acquisition.

In the present study, however, real-time imaging was performed and the images were available online without delay. This availability allowed on-the-flight planning of the different imaging planes during real-time imaging, which in turn kept the overall time to perform one real-time sequence shorter than 5 minutes and thereby made this examination a potentially cost-effective and practical approach for routine application.

Schalla et al (26) recently compared real-time gradient-echo cardiac MR imaging with conventional breath-hold gradient-echo MR imaging for regional wall motion assessment during dobutamine-induced stress. Excellent agreement between the two sequences was reported. Their system, similar to that in the Plein et al study, also did not allow online reconstruction of real-time MR images, and the spatial resolution was lower than that achieved with the sequences used in the current study.

A limitation of the present study is that the observers were not blinded with regard to the different cardiac MR imaging sequences performed, so the modalities used were easily identified. This might have affected the evaluation of the two real-time MR imaging sequences. Analysis of the images acquired with the reference method (ie, BH-SSFP) was performed at a different workstation at least 1 week after analysis of the real-time images obtained in each patient. Therefore, we believe that our study results were not substantially influenced by the lack of blinding.

In conclusion, interactive radial steady-state free precession cardiac MR imaging enables online assessment of regional myocardial function in real time in only 5 minutes, with results that are in high agreement with those of conventional breath-hold cine cardiac MR imaging. The high performance of radial MR imaging is attributed to the excellent image quality achieved by combining radial imaging with steady-state free precession imaging. Moreover, the radial sequence is superior to the spiral sequence and ECHO for the visualization of myocardial wall segments. The radial sequence should be the preferred real-time cardiac MR imaging sequence for assessment of segmental wall motion in patients with nondiagnostic ECHO results.


   FOOTNOTES
 
Abbreviation: BH-SSFP = breath-hold steady-state free precession, ECHO = echocardiography

Author contributions: Guarantors of integrity of entire study, H.P.K., A.B.; study concepts, H.P.K., A.B., A.F.; study design, H.P.K., E.S., A.B., A.F.; literature research, A.W.; clinical studies, H.P.K., E.S., J.S., A.W.; data acquisition, E.S., A.W.; data analysis/interpretation, H.P.K., J.S., E.S., A.W., A.B., A.F.; statistical analysis, H.P.K., N.H.; manuscript preparation, H.P.K., E.S.; manuscript definition of intellectual content, H.P.K., A.B., A.F.; manuscript editing, H.P.K., E.S.; manuscript revision/review, A.B., A.F.; manuscript final version approval, R.W.G., P.H.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 

  1. Rubin DN, Yazbek N, Garcia MJ, Stewart WJ, Thomas JD. Qualitative and quantitative effects of harmonic echocardiographic imaging on endocardial edge definition and side-lobe artifacts. J Am Soc Echocardiogr 2000; 13:1012-1018.[CrossRef][Medline]
  2. Senior R, Soman P, Khattar RS, Lahiri A. Improved endocardial visualization with second harmonic imaging compared with fundamental two-dimensional echocardiographic imaging. Am Heart J 1999; 138:163-168.[CrossRef][Medline]
  3. Spencer KT, Bednarz J, Rafter PG, Korcarz C, Lang RM. Use of harmonic imaging without echocardiographic contrast to improve two-dimensional image quality. Am J Cardiol 1998; 82:794-799.[CrossRef][Medline]
  4. Malhotra V, Nwogu J, Bondmass MD, et al. Is the technically limited echocardiographic study an endangered species? Endocardial border definition with native tissue harmonic imaging and Optison contrast—a review of 200 cases. J Am Soc Echocardiogr 2000; 13:771-773.[CrossRef][Medline]
  5. Kasprzak JD, Paelinck B, Ten Cate FJ, et al. Comparison of native and contrast-enhanced harmonic echocardiography for visualization of left ventricular endocardial border. Am J Cardiol 1999; 83:211-217.[Medline]
  6. Spencer KT, Bednarz J, Mor-Avi V, et al. The role of echocardiographic harmonic imaging and contrast enhancement for improvement of endocardial border delineation. J Am Soc Echocardiogr 2000; 13:131-138.[Medline]
  7. Weber OM, Eggers H, Spiegel MA, Scheidegger MB, Proksa R, Boesiger P. Real-time interactive magnetic resonance imaging with multiple coils for the assessment of left ventricular function. J Magn Reson Imaging 1999; 10:826-832.[CrossRef][Medline]
  8. Setser RM, Fischer SE, Lorenz CH. Quantification of left ventricular function with magnetic resonance images acquired in real time. J Magn Reson Imaging 2000; 12:430-438.[CrossRef][Medline]
  9. Nagel E, Schneider U, Schalla S, et al. Magnetic resonance real-time imaging for the evaluation of left ventricular function. J Cardiovasc Magn Reson 2000; 2:7-14.[Medline]
  10. Schalla S, Nagel E, Lehmkuhl H, et al. Comparison of magnetic resonance real-time imaging of left ventricular function with conventional magnetic resonance imaging and echocardiography. Am J Cardiol 2001; 87:95-99.[CrossRef][Medline]
  11. Plein S, Bloomer TN, Ridgway JP, Jones TR, Bainbridge GJ, Sivananthan MU. Steady-state free precession magnetic resonance imaging of the heart: comparison with segmented k-space gradient-echo imaging. J Magn Reson Imaging 2001; 14:230-236.[CrossRef][Medline]
  12. Yang PC, Kerr AB, Liu AC, et al. New real-time interactive cardiac magnetic resonance imaging system complements echocardiography. J Am Coll Cardiol 1998; 32:2049-2056.[Abstract/Free Full Text]
  13. Barkhausen J, Ruehm SG, Goyen M, Buck T, Laub G, Debatin JF. MR evaluation of ventricular function: true fast imaging with steady-state precession versus fast low-angle shot cine MR imaging—feasibility study. Radiology 2001; 219:264- 269.[Abstract/Free Full Text]
  14. Schaeffter T, Weiss S, Eggers H, Rasche V. Projection reconstruction balanced fast field echo for interactive real-time cardiac imaging. Magn Reson Med 2001; 46:1238-1241.[CrossRef][Medline]
  15. Shankaranarayanan A, Simonetti OP, Laub G, Lewin JS, Duerk JL. Segmented k-space and real-time cardiac cine MR imaging with radial trajectories. Radiology 2001; 221:827-836.[Abstract/Free Full Text]
  16. Spuentrup E, Kuehl HP, Wall A, Heer C, Buecker A. Visualization of cardiac myxoma mobility with real-time spiral magnetic resonance imaging. Circulation 2001; 104:E101.
  17. Riederer SJ, Tasciyan T, Farzaneh F, Lee JN, Wright RC, Herfkens RJ. MR fluoroscopy: technical feasibility. Magn Reson Med 1988; 8:1-15.[Medline]
  18. Eggers H, Proska R. Multiprocessor system for real-time convolution interpolation reconstruction (abstr) Proceedings of the Seventh Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 1999; 95.
  19. Schiller NB, Shah PM, Crawford M, et al. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography: American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-Dimensional Echocardiograms. J Am Soc Echocardiogr 1989; 2:358- 367.[Medline]
  20. Franke A, Hoffmann R, Kühl HP, et al. Non-contrast second harmonic imaging improves interobserver agreement and accuracy of dobutamine stress echocardiography in patients with impaired image quality. Heart 2000; 83:133-140.[Abstract/Free Full Text]
  21. Fang W, Peerless FS, Bundy J. Evaluating left ventricular function using real-time true FISP: a comparison with conventional MR techniques (abstr) Proceedings of the Eighth Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2000; 308.
  22. Chien D, Edelman RR. Ultrafast imaging using gradient echoes. Magn Reson Q 1991; 7:31-56.[Medline]
  23. Wiesmann F, Gatehouse PD, Panting JR, Taylor AM, Firmin DN, Pennell DJ. Comparison of fast spiral, echo planar, and fast low-angle shot MRI for cardiac volumetry at .5T. J Magn Reson Imaging 1998; 8:1033-1039.[Medline]
  24. Atalay MK, Poncelet BP, Kantor HL, Brady TJ, Weisskoff RM. Cardiac susceptibility artifacts arising from the heart-lung interface. Magn Reson Med 2001; 45:341-345.[CrossRef][Medline]
  25. Plein S, Smith WH, Ridgway JP, et al. Qualitative and quantitative analysis of regional left ventricular wall dynamics using real-time magnetic resonance imaging: comparison with conventional breath-hold gradient echo acquisition in volunteers and patients. J Magn Reson Imaging 2001; 14:23-30.[CrossRef][Medline]
  26. Schalla S, Klein C, Paetsch I, et al. Real-time MR image acquisition during high-dose dobutamine hydrochloride stress for detecting left ventricular wall-motion abnormalities in patients with coronary arterial disease. Radiology 2002; 224:845-851.[Abstract/Free Full Text]
  27. Peters DC, Korosec FR, Grist TM, et al. Undersampled projection reconstruction applied to MR angiography. Magn Reson Med 2000; 43:91-101.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 RadiologyHome page
V. Muthurangu, P. Lurz, J. D. Critchely, J. E. Deanfield, A. M. Taylor, and M. S. Hansen
Real-time Assessment of Right and Left Ventricular Volumes and Function in Patients with Congenital Heart Disease by Using High Spatiotemporal Resolution Radial k-t SENSE
Radiology, September 1, 2008; 248(3): 782 - 791.
[Abstract] [Full Text] [PDF]

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

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