Acute Myocarditis: Diagnostic Value of Contrast-Enhanced Cine Steady-State Free Precession MRI Sequences

Transcription

1 Cardiopulmonary Imaging Original Research Cardiopulmonary Imaging Original Research Jean-François Deux 1,2 Mezri Maatouk 1 Pascal Lim 3 Alexandre Vignaud 4 Julie Mayer 1 Pascal Gueret 3 Alain Rahmouni 1 Deux JF, Maatouk M, Lim P, et al. Keywords: acute myocarditis, cine MRI, late gadolinium enhancement, MRI DOI:1.2214/AJR Received October 29, 21; accepted after revision March 2, Radiology Department, Henri Mondor Hospital, Assistance Publique-Hôpitaux de Paris, University Paris Est, Créteil, F-941, France. Address correspondence to J. F. Deux (jean-francois.deux@hmn.aphp.fr). 2 CNRS UMR 754, Centre de Recherches Chirurgicales, Henri Mondor Hospital, Assistance Publique-Hôpitaux de Paris, University Paris Est, Créteil, France. 3 Cardiology Department, Henri Mondor Hospital, Assistance Publique-Hôpitaux de Paris, University Paris Est, Créteil, France. 4 Siemens Healthcare, Saint Denis, France. AJR 211; 197: X/11/ American Roentgen Ray Society Acute Myocarditis: Diagnostic Value of Contrast-Enhanced Cine Steady-State Free Precession MRI Sequences OBJECTIVE. MRI has become the primary tool for assessment of myocardial inflammation in patients with suspected acute myocarditis. Optimal diagnostic performance is achieved with late gadolinium-enhanced sequences, but cine balanced steady-state free precession (SSFP) MRI sequences are routinely used to evaluate cardiac function. Our aim was to prospectively assess the diagnostic value of unenhanced and contrast-enhanced cine SSFP MRI sequences in comparison with late gadolinium-enhanced sequences for imaging of patients with strong evidence of acute myocarditis. SUBJECTS AND METHODS. Eighteen patients with strong evidence of acute myocarditis underwent 1.5-T cardiac MRI. Unenhanced and contrast-enhanced cine SSFP images and late gadolinium-enhanced images were obtained. The images were analyzed both qualitatively and quantitatively. Data were analyzed with analysis of variance and the Bonferroni test or paired Student t test. RESULTS. Areas of high signal intensity were detected in 28% (5/18), 94% (17/18), and 89% (16/18) of patients on unenhanced cine, contrast-enhanced cine, and late gadoliniumenhanced images. In one patient, contrast-enhanced cine images revealed subepicardial areas of high signal intensity that were not visible on late gadolinium-enhanced images. The location and transmural nature of involved segments matched on contrast-enhanced cine and late gadolinium-enhanced images (both, r =.91, p <.1). The contrast-to-noise ratio was significantly higher on contrast-enhanced cine images than on late gadolinium-enhanced images (p <.5). CONCLUSION. Contrast-enhanced cine MRI is a valuable tool for detection of lesions of acute myocarditis and should be recommended for routine clinical MRI. C ardiac MRI has become the primary tool for the assessment of myocardial inflammation in patients with suspected acute myocarditis [1]. Late gadolinium-enhanced sequences are considered key for detecting lesions of acute myocarditis characterized by subepicardial enhancement foci and sparing of the subendocardial layers [2, 3]. Cine balanced steady-state free precession (SSFP) MRI sequences, which yield images with high contrast between blood and myocardium, are routinely used before contrast administration to evaluate cardiac function [4]. Owing to the T1 component of the signal, SSFP MRI sequences after gadolinium administration have also been positively evaluated in comparison with late gadolinium-enhanced sequences in imaging of patients with myocardial infarction [5, 6] and of patients with acute coronary syndrome [7]. The signal resulting from SSFP sequences depends on the T2/T1 tissue relaxation time ratio according to the following simplified formula [8]: In areas of myocardial inflammation, the increased T2 relaxation time due to edema combined with the decreased T1 relaxation time after gadolinium administration due to local gadolinium concentration increases the signal intensity in contrast-enhanced SSFP MRI sequences, allowing detection of myocardial lesions. To our knowledge, the value of contrast-enhanced cine SSFP sequences in the detection of foci of acute myocarditis has not been studied prospectively. The aim of this study was to compare the efficiency of con- AJR:197, November

2 trast-enhanced cine and late gadolinium-enhanced sequences in imaging of patients with strong evidence of acute myocarditis. Subjects and Methods This prospective study was approved by the institutional review board of our hospital, and informed consent was obtained from all patients. Patient Sample Between November 28 and December 29, 22 consecutively registered patients with strong evidence of acute myocarditis were included in this prospective study. The inclusion criteria were as follows: symptoms of heart disease (acute chest pain, dyspnea, or palpitations), evidence of recent myocardial injury (ECG changes, elevated troponin T concentration, or ventricular dysfunction at echocardiography), recent history of infection, and absence of coronary artery disease at coronary angiography. All coronary arteriograms were analyzed in consensus by two experienced observers. A normal coronary angiographic finding was defined as absence of stenosis and regularity of coronary vessels. Four patients were excluded because of refusal to undergo MRI (one patient), contraindication to MRI (one patient), and creatinine clearance less than 3 ml/min (two patients). The other 18 patients (15 men, three women; mean age, 38.2 ± 14 [SD] years; median, 37.5 years; range, years) were included in the study. The mean delay between onset of clinical symptoms and cardiac MRI was 5.2 ± 3 days (median, 4.5 days; range, 1 1 days). MRI Cardiac MRI was performed with a 1.5-T system (Magnetom Avanto, Siemens Healthcare) equipped with a high-performance gradient subsystem (maximum amplitude, 4 mt/m; minimum rise, 2 µs) and an 8-channel phased-array cardiac coil. Gradient-echo images were first used to locate the true anatomic axes of the heart. Unenhanced cine SSFP sequences were performed for each patient. Five short-axis sections encompassing the left ventricle, one four-chamber section, and one two-chamber long-axis section were acquired. The following parameters were used: TR/TE, 2.8/1.4 (apparent TR, 31.4 ms; 11 segments); flip angle, 82 ; matrix size, ; FOV, 3 27 mm; slice thickness, 6 mm. Retrospective ECG gating was used with 2 phases per section. Parallel imaging was performed by generalized autocalibrating partial parallel acquisition algorithm with an acceleration factor limited to 2 to reduce the spatial dependence on noise on accelerated images [9]. Image acquisition lasted 1 15 seconds, depending on the heart rate. Five minutes after IV administration of a bolus of gadolinium chelate (gadoterate dimeglumine, Dotarem, Guerbet;.2 mmol/kg body weight, injected at a rate of 3.5 ml/s, followed by a 3-mL saline flush), contrast-enhanced cine images were obtained for all patients with the same parameters and in the same location as for the unenhanced cine images. Ten minutes after the beginning of the contrast injection, 3D late gadolinium-enhanced acquisition was performed for all patients with a segmented 3D inversion recovery gradientecho T1-weighted technique. Acquisition was performed during the diastolic phase. An inversion recovery time scouting sequence was systematically used to adjust the optimal inversion recovery time. The sequence parameters were as follows: TR/TE, 3.9/1.4; mean inversion recovery time, 22 ± 5 ms; flip angle, 1 ; matrix size, ; FOV, 3 27 mm; number of sections, 12; slice thickness, 6 mm. Image acquisition lasted 12 2 seconds, depending on heart rate. Two complete sets TABLE 1: Characteristics of Study Sample Patient No. Sex Age (y) Symptom of contiguous short-axis sections (24 sections), one set of four-chamber sections, and one set of twochamber long-axis sections (encompassing the entire left ventricle) were acquired. MRI Analysis All images were analyzed anonymously with an acquisition platform (Leonardo, Siemens Healthcare) specifically for this study. Qualitative analysis Qualitative interpretation of images from each MRI procedure was performed separately in two consecutive sessions 2 months apart by two MRI readers with 8 and 3 years of experience in cardiac MRI. The readers were blinded to the clinical data and the results at the other reading session. In each session, unenhanced cine images, contrast-enhanced cine images, and late gadolinium-enhanced images were analyzed independently by the readers. An overall image quality score based on the presence of artifacts linked to poor breath-hold and quality of signal nulling of the myocardium on late gad- Left Ventricular Ejection Fraction (%) a Troponin T Concentration (ng/ml) b ECG Finding 1 M 32 Acute chest pain 51 ST wave M 38 Acute chest pain 63 ST wave M 59 Acute chest pain 51 ST wave 5. 4 F 72 Acute chest pain 36 Normal M 36 Palpitation 44 Negative T 3. wave 6 M 24 Acute chest pain 59 T wave M 29 Acute chest pain 4 ST wave M 41 Acute chest pain 54 Normal M 29 Acute chest pain 48 ST wave M 19 Palpitation 54 ST wave F 43 Acute chest pain 42 ST wave M 24 Dyspnea 6 ST wave M 37 Acute chest pain 5 ST wave.9 14 M 18 Acute chest pain 6 ST wave F 41 Acute chest pain 5 Normal M 51 Acute chest pain 6 Normal M 56 Acute chest pain 66 ST wave M 38 Palpitation 53 ST wave 6.4 Mean Median 37.5 SD Note All patients had normal coronary angiographic findings. a Calculated at echocardiography. b Measurement by immunoassay of serum samples; upper limit < AJR:197, November 211

3 olinium-enhanced images was given to each set of images: 1, poor; 2, medium; 3, good; 4, excellent. The presence, location, and extent of areas of high signal intensity based on the 17-segment American Heart Association classification [1] were noted for each image. The location of high-signal-intensity foci in the left ventricular wall was noted as subendocardial, midwall, subepicardial, or transmural. Quantitative analysis The two readers determined by consensus a reference image plane in which the high-signal-intensity areas were most visible on late gadolinium-enhanced images (a short-axis section and a four-chamber or two-chamber long-axis section). In this reference plane, both the entire left myocardium and highsignal-intensity areas detected in the diastolic and systolic phases of unenhanced cine imaging, in the diastolic and systolic phases of contrast-enhanced cine imaging, and during late gadoliniumenhanced imaging were manually contoured. The mean signal intensity (SI high-signal foci ± SD) and absolute surface area of high signal intensity were noted. The epicardial and endocardial wall borders of the left ventricle were manually contoured, and the myocardial area was measured. The relative surface area of high signal intensity present in the myocardium was defined as follows: relative surface area of high signal intensity = (relative surface area of high signal intensity / surface area of left myocardium) 1. The signal intensities of normal myocardium (SI normal myocardium ) and background air (SI noise ) were determined by placement of 1-cm 2 regions of interest in the myocardium avoiding areas of high signal intensity (SI normal myocardium ) and within background air (SI noise ). Blood-pool signal intensity (SI blood pool ) was calculated by placement of a region of interest encompassing the left ventricular cavity. Signal-to-noise ratio (SNR) and contrastto-noise ratio (CNR) were calculated as follows: SNR = SI high-signal foci / SD of SI noise CNR = (SI high-signal foci SI normal myocardium ) / SD of SI noise The CNR of the blood pool was calculated as follows: CNR blood pool = (SI high-signal foci SI normal myocardium ) / SD of SI blood pool Last, the contrast difference index was calculated as follows: Contrast difference index = (SI high-signal foci SI normal myocardium ) / SD of SI normal myocardium Statistical Analysis Data were presented as mean ± SD. The significance of differences was analyzed with analysis of variance and Bonferroni test because the Kolmogorov-Smirnov test and histograms (i.e., Q-Q plots) revealed normal distribution of the data. A nonparametric Wilcoxon test was used to compare sequences with respect to image quality. A value of p <.5 was considered significant. Spearman correlation was conducted on the different sequences for measurement of the surface areas of high signal intensity, for location, and for transmural nature of lesions. The Cohen kappa coefficient for determining intraobserver agreement in detecting areas of high signal intensity was calculated for unenhanced cine sequences, contrastenhanced cine sequences, and late gadoliniumenhanced sequences. The SPSS program (version 16., SPSS) was used for all analyses. Results Table 1 shows the characteristics of the study sample. Acute chest pain was the most frequent symptom, mentioned by 78% (14/18) of patients. Electrical abnormalities were detected in 78% of cases (14/18) and usually consisted of ST-segment elevation. Qualitative Analysis The mean image quality score was 3 or 4 for the unenhanced cine images of 1% of the patients (n = 18), for the contrast-enhanced cine images of 89% of the patients (n = 16), and the late gadolinium-enhanced images of 89% of the patients (n = 16). The mean image quality score was 2 for both the Fig year-old man (patient 13) with acute chest pain, elevated troponin concentration, and normal coronary angiographic findings undergoing evaluation for acute myocarditis. Surface area (S) of lesions of acute myocarditis is greater on systolic contrast-enhanced cine image (C) than on late gadolinium-enhanced image (D). Surface area of lesions on diastolic contrast-enhanced and late gadolinium-enhanced images are in same range. Contrast-to-noise ratio (CNR) of contrast-enhanced cine images (B and C) is higher than CNR of late gadolinium-enhanced image. A, Unenhanced cine MR image in four-chamber view shows no lesion in left myocardium. B and C, Diastolic (B) and systolic (C) contrast-enhanced cine MR images acquired in same plane show foci of high signal intensity (arrows) in midportion of left lateral wall. D, Late gadolinium-enhanced MR image shows focal enhancement (arrows) in same areas as in B and C. AJR:197, November

4 contrast-enhanced cine and late gadoliniumenhanced images of 11% of the patients (n = 2). There were no statistical differences between sequences. Areas of high signal intensity were detected in 28% (5/18), 94% (17/18), and 89% (16/18) of the patients on unenhanced cine, contrast-enhanced cine, and late gadolinium-enhanced sequences. All 16 patients with areas of high signal intensity in the myocardium on late gadolinium-enhanced images also had areas of high signal intensity in the same locations within the myocardium on contrast-enhanced cine images (Fig. 1). High-signal-intensity foci were located in the subepicardial part of the left ventricular wall in 63% of the patients (1/16). They were located in the midwall in 31% of patients (5/16) and were transmural in 6% of patients (1/16). High-signal-intensity foci were never limited to the subendocardial area. The contrast-enhanced cine images of one patient (patient 12) revealed subepicardial high-signal-intensity areas in the lateral part of the left ventricular wall (segments 1 and 11) that were not visible on late gadoliniumenhanced images (Fig. 2). One patient (patient 14) had normal patterns on images obtained with all sequences. The mean number of affected segments was 3.6 ± 1.8 (range, 1 7) on late gadolinium-enhanced images and 3.5 ± 1.8 (range, 1 7) on contrast-enhanced cine images. The signal foci in the left ventricular wall on the contrast-enhanced images correlated closely with those on the late gadolinium-enhanced images in terms of affected segments and transmural nature of the lesion (both, Fig year-old man (patient 12) with dyspnea and electrical abnormalities. Foci of subepicardial areas of high signal intensity suggesting lesions of acute myocarditis are evident on contrast-enhanced cine images but not on late gadolinium-enhanced image. A, Unenhanced cine MR image acquired in short-axis view shows no abnormal findings. B and C, Diastolic (B) and systolic (C) contrast-enhanced cine MR images show enhancement (arrowheads) in subepicardial area of left lateral free wall. D, Late gadolinium-enhanced image shows no abnormal findings. r =.91, p <.1). Lateral segments were most likely to be affected: the laterobasal, lateromedial, and lateroapical segments were affected in 5% (9/18), 56% (1/18), and 67% (12/18) of the patients. The inferobasal, inferomedial, and inferoapical segments were affected in 28% (5/18), 17% (3/18), and 11% (2/18) of the patients. The anterior and septal segments were affected in 22% (4/18) and 11% (2/18) of the patients. Quantitative Analysis The surface area, SNR, and CNR of highsignal-intensity areas are shown in Tables 2 and 3. In the 16 patients who had abnormal findings on both contrast-enhanced cine and late gadolinium-enhanced images, the surface areas of high signal intensity measured in the diastolic phase of the contrast-enhanced cine sequence correlated with those measured on late gadolinium-enhanced images (absolute, r =.89, p <.1; relative, r =.84, p <.1). In addition, the surface areas of high signal intensity measured in the systolic phase of the contrast-enhanced cine sequence and the late gadolinium-enhanced sequence also correlated closely (absolute and relative, r =.82, p <.1). The results of linear regression and the Bland-Altman plots of the comparison of absolute surface areas on the late gadolinium-enhanced and contrast-enhanced cine images are shown in Figure 3. The surface areas of high signal intensity measured in the systolic phase of the contrast-enhanced cine sequence were larger than those measured in the diastolic phase (absolute, p <.1; relative, p <.5) and those measured on late gadolinium-enhanced images (absolute, p <.5; relative, p <.5). SNR and CNR were significantly higher on contrast-enhanced cine images than on late gadolinium-enhanced images (SNR, p <.1; CNR, p <.5). During cine sequences, both SNR and CNR increased after gadolinium administration. The blood pool CNR was significantly (p <.1) higher on contrast-enhanced cine images (diastolic, 11.9 ± 4; systolic, 11.3 ± 3) than on late gadolinium-enhanced images (7.1 ± 5). The contrast difference index was in the same range (p not significant) for the contrast-enhanced cine sequences (diastolic, 5.9 ± 2; systolic, 6.1 ± 3) and the late gadolinium-enhanced sequence (6.4 ± 3). The Cohen kappa coefficient for intraobserver agreement in detecting areas of high signal intensity was.65 in the diastolic phase of the unenhanced cine sequences and 184 AJR:197, November 211

5 TABLE 2: Absolute and Mean Relative Surface Areas Measured on Unenhanced Cine, Contrast-Enhanced Cine, and Late Contrast- Enhanced Images Area (cm 2 ) Unenhanced Cine Contrast-Enhanced Cine Patient No. Diastolic Systolic Diastolic Systolic Late Contrast-Enhanced Mean ± SD.15 ±.4.46 ± ± ± ± 1.4 Relative surface area in the systolic phase;.7 in the diastolic phase of contrast-enhanced cine sequences and.8 in the systolic phase; and.8 in the late gadolinium-enhanced sequences. Discussion In this study, we found that the diagnostic value of contrast-enhanced cine sequences in the detection of foci of acute myocarditis was as high as that of late gadolinium-enhanced sequences in imaging of 18 patients with suspected acute myocarditis. Our results suggest that the contrast-enhanced cine sequence may be proposed as an alternative for detecting lesions of acute myocarditis. Cine balanced SSFP sequences are commonly used to analyze local and global cardiac function before gadolinium administration [11, 12]. Because T1 and T2 contrast is intrinsically present on the images, a cine sequence can be used after gadolinium administration to detect areas of enhancement within the myocardium. In the detection and location of lesions of myocardial infarction, Laissy et al. [5] found good correlation between late gadolinium-enhanced and contrast-enhanced cine images. Another advantage is that contrast injection does not preclude measurement of left ventricular function on contrast-enhanced cine images [13]. Codreanu et al. [7] reported that contrastenhanced cine images can depict lesions of acute myocarditis or myocardial infarction in acute coronary syndrome. To the best of our knowledge, contrast-enhanced cine sequences have never been systematically evaluated in the detection of lesions in patients with strong evidence of acute myocarditis with late gadolinium-enhanced imaging as the standard of reference. In this study, we found that a contrast-enhanced cine sequence can be used to detect the lesions of acute myocarditis rapidly and efficiently, a few minutes after injection, in comparison with use of the standard late gadolinium-enhanced sequence. Despite a difference in acquisition time between sequences (5 and 1 minutes for the contrast-enhanced cine and late gadolinium-enhanced sequences), we found good correlation, first, between the surface areas measured in the late gadolinium-enhanced and the diastolic phases of the contrast-enhanced cine sequence and, second, between the locations of abnormalities. Our results are in agreement with those of previous studies of the use of contrast-enhanced cine sequences in the diagnosis of myocardial infarction [5, 6]. The surface areas of lesions measured in the systolic phase of contrast-enhanced cine imaging was greater than that measured in the diastolic phase (either late gadoliniumenhanced or diastolic contrast-enhanced cine images). We hypothesize that stretching of the lesions of acute myocarditis during the systolic phase can explain this finding, which can be used to increase diagnostic confidence that a lesion is acute myocarditis when the findings on diastolic late gadolinium-enhanced images are doubtful, especially in cases of inadequate determination of inversion time. An interesting finding in this study was that the SNR and CNR of acute myocarditis lesions were significantly higher on contrastenhanced cine images than on late gadolinium-enhanced images. In a previous study, Laissy et al. [5] found similar enhancement ratios both for contrast-enhanced cine and late gadolinium-enhanced sequences. Although cine sequences are known to result in a high SNR and CNR, these results may be surprising because nullification of the normal myocardial signal on late gadoliniumenhanced images improves the CNR. It is noteworthy that the CNR of late gadoliniumenhanced sequences should be equal to the SNR if the nullification of healthy myocardium is perfect (i.e., equal to ). Nevertheless, in clinical practice, signal nullification of normal myocardium on late gadoliniumenhanced images may be suboptimal, leading to differences between CNR and SNR. In our study, the mean CNR (17.1) was lower than the mean SNR (25.5) on late gadolinium-enhanced images, reflecting that the mean signal intensity of normal myocardium was greater than. In the current study, late gadolinium-enhanced images depicted areas of contrast enhancement in 89% of patients. This proportion is higher than the 44% sensitivity reported by Abdel-Aty et al. [14] for spin-echo T1-weighted sequences and the 52% report- AJR:197, November

6 TABLE 3: Signal- and Contrast-to-Noise Ratios for Unenhanced Cine, Contrast-Enhanced Cine, and Late Contrast-Enhanced Images Late Contrast- Unenhanced Cine MRI Contrast-Enhanced Cine MRI Enhanced Patient Diastolic Systolic Diastolic Systolic No. SNR CNR SNR CNR SNR CNR SNR CNR SNR CNR NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS 13 NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS 15 NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS NAS Mean SD Note SNR = signal-to-noise ratio, CNR = contrast-to-noise ratio, NAS = no abnormality seen in signal intensity. Systolic Contrast- Enhanced Cine Images Diastolic Contrast- Enhanced Cine Images r =.82 p < Late Gadolinium-Enhanced Images r =.89 p < Late Gadolinium-Enhanced Images Difference Difference SD mean 2 SD Mean SD mean 2 SD Mean ed by Gahide et al. [2] for late gadoliniumenhanced sequences but is in agreement with the 88% sensitivity reported by Mahrholdt et al. [15] for gradient-echo inversion recovery T1-weighted sequences. The locations of abnormalities, predominantly in the epicardial portion of the lateral free wall, are also in agreement with those in previous studies [13, 16, 17]. One additional case of subepicardial areas of high signal intensity suggestive of lesions of acute myocarditis was detected on contrast-enhanced cine images but not identified as such on late gadolinium-enhanced images. Because of the lack of endomyocardial biopsy to confirm the diagnosis of acute myocarditis, it is difficult to know whether this case represents a false-negative finding of late gadolinium-enhanced imaging or a false-positive finding of contrast-enhanced cine imaging. A T2-weighted sequence might have been useful in this case to detect myocardial edema, which in rare cases of acute myocarditis can be seen only on T2- weighted images when the findings are normal with a late enhancement sequence [1]. A limitation of this study was that the final diagnosis of acute myocarditis was based on a combination of clinical, electrical, biologic, and angiographic findings. Endomyocardial biopsy was not performed. Confirmatory histopathologic, immunohistologic, and virus polymerase chain reaction data therefore were unavailable. We rarely perform endomyocardial biopsy because we consider it excessively invasive for patients with a mild condition that has a generally favorable outcome. Another limitation of our study was that we performed contrast-enhanced cine sequences a few minutes before late gadolinium-enhanced sequences and thus could not strictly compare images obtained with these sequences. Last, diffuse myocardial changes were not analyzed in our patient sample, although such changes have been documented in cases of acute myocarditis [18]. Fig. 3 Graphs show linear regression (left) and Bland-Altman (right) comparisons of absolute surface area (square centimeters) of lesions on late gadolinium-enhanced and systolic (top) and diastolic (bottom) contrast-enhanced cine images. Bland- Altman graphs show mean surface area on contrastenhanced cine images and late gadolinium-enhanced images plotted against difference between the two. Solid line indicates mean of differences; dashed lines, limits of agreement. 186 AJR:197, November 211

7 References 1. Friedrich MG, Sechtem U, Schulz-Menger J, et al. Cardiovascular magnetic resonance in myocarditis: a JACC white paper. J Am Coll Cardiol 29; 53: Gahide G, Bertrand D, Roubille F, et al. MR delayed enhancement imaging findings in suspected acute myocarditis. Eur Radiol 21; 2: Mahrholdt H, Wagner A, Deluigi CC, et al. Presentation, patterns of myocardial damage, and clinical course of viral myocarditis. Circulation 26; 114: Chavhan GB, Babyn PS, Jankharia BG, Cheng HL, Shroff MM. Steady-state MR imaging sequences: physics, classification, and clinical applications. RadioGraphics 28; 28: Laissy JP, Hyafil F, Huart V, et al. Value of contrast-enhanced, balanced cine-mr sequences in the assessment of apparent infarct size after acute myocardial infarction: a prospective comparison with delayed-enhancement sequences. J Magn Reson Imaging 25; 22: Setser RM, Kim JK, Chung YC, et al. Cine delayed-enhancement MR imaging of the heart: initial experience. Radiology 26; 239: Codreanu A, Djaballah W, Angioi M, et al. Detection of myocarditis by contrast-enhanced MRI in patients presenting with acute coronary syndrome but no coronary stenosis. J Magn Reson Imaging 27; 25: Bernstein MA. Gradient echo. In: Bernstein MA, King KF, Zhou XJ. Handbook of MRI pulse sequences. Boston, MA: Academic Press, 24: Reeder SB, Wintersperger BJ, Dietrich O, et al. Practical approaches to the evaluation of signal-tonoise ratio performance with parallel imaging: application with cardiac imaging and a 32-channel cardiac coil. Magn Reson Med 25; 54: Cerqueira MD, Weissman NJ, Dilsizian V, et al. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation 22; 15: Hudsmith LE, Petersen SE, Francis JM, Robson MD, Neubauer S. Normal human left and right ventricular and left atrial dimensions using steady state free precession magnetic resonance imaging. J Cardiovasc Magn Reson 25; 7: Malayeri AA, Johnson WC, Macedo R, Bathon J, Lima JA, Bluemke DA. Cardiac cine MRI: quantification of the relationship between fast gradient echo and steady-state free precession for determination of myocardial mass and volumes. J Magn Reson Imaging 28; 28: Lasalarie JC, Serfaty JM, Carre C, et al. Accuracy of contrast-enhanced cine-mr sequences in the assessment of left ventricular function. Eur Radiol 27; 17: Abdel-Aty H, Boye P, Zagrosek A, et al. Diagnostic performance of cardiovascular magnetic resonance in patients with suspected acute myocarditis: comparison of different approaches. J Am Coll Cardiol 25; 45: Mahrholdt H, Goedecke C, Wagner A, et al. Cardiovascular magnetic resonance assessment of human myocarditis: a comparison to histology and molecular pathology. Circulation 24; 19: Friedrich MG, Strohm O, Schulz-Menger J, Marciniak H, Luft FC, Dietz R. Contrast media-enhanced magnetic resonance imaging visualizes myocardial changes in the course of viral myocarditis. Circulation 1998; 97: Laissy JP, Hyafil F, Feldman LJ, et al. Differentiating acute myocardial infarction from myocarditis: diagnostic value of early- and delayed-perfusion cardiac MR imaging. Radiology 25; 237: Laissy JP, Messin B, Varenne O, et al. MRI of acute myocarditis: a comprehensive approach based on various imaging sequences. Chest 22; 122: AJR:197, November