Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Cardiovascular disease wikipedia , lookup
History of invasive and interventional cardiology wikipedia , lookup
Remote ischemic conditioning wikipedia , lookup
Antihypertensive drug wikipedia , lookup
Arrhythmogenic right ventricular dysplasia wikipedia , lookup
Drug-eluting stent wikipedia , lookup
Jatene procedure wikipedia , lookup
Quantium Medical Cardiac Output wikipedia , lookup
Bright-Blood T2-Weighted MRI Has Higher Diagnostic Accuracy Than Dark-Blood Short Tau Inversion Recovery MRI for Detection of Acute Myocardial Infarction and for Assessment of the Ischemic Area at Risk and Myocardial Salvage Alexander R. Payne, MBChB; Matthew Casey, BMedSci; John McClure, PhD; Ross McGeoch, MBChB; Aengus Murphy, MBChB; Rosemary Woodward, BSc; Andrew Saul, BSc; Xiaoming Bi, PhD; Sven Zuehlsdorff, PhD; Keith G. Oldroyd, MD(Hons), FRCP, FSCAI; Niko Tzemos, MD(Hons), FRCP, FASE; Colin Berry, BSc, PhD, FRCP, FACC Downloaded from http://circimaging.ahajournals.org/ by guest on May 11, 2017 Background—T2-Weighted MRI reveals myocardial edema and enables estimation of the ischemic area at risk and myocardial salvage in patients with acute myocardial infarction (MI). We compared the diagnostic accuracy of a new bright-blood T2-weighted with a standard black blood T2-weighted MRI in patients with acute MI. Methods and Results—A breath-hold, bright-blood T2-weighted, Acquisition for Cardiac Unified T2 Edema pulse sequence with normalization for coil sensitivity and a breath-hold T2 dark-blood short tau inversion recovery sequence were used to depict the area at risk in 54 consecutive acute MI patients. Infarct size was measured on gadolinium late contrast enhancement images. Compared with dark-blood T2-weighted MRI, consensus agreements between independent observers for identification of myocardial edema were higher with bright-blood T2-weighted MRI when evaluated per patient (P⬍0.001) and per segment of left ventricle (P⬍0.001). Compared with bright-blood T2-weighted MRI, dark-blood T2-weighted MRI underestimated the area at risk compared with infarct size (P⬍0.001). The 95% limits of agreement for interobserver agreements for the ischemic area at risk and myocardial salvage were wider with dark-blood T2-weighted MRI than with bright-blood T2-weighted MRI. Bright blood enabled more accurate identification of the culprit coronary artery with correct identification in 94% of cases compared with 61% for dark blood (P⬍0.001). Conclusions—Bright-blood T2-weighted MRI has higher diagnostic accuracy than dark-blood T2-weighted MRI. Additionally, dark-blood T2-weighted MRI may underestimate area at risk and myocardial salvage. (Circ Cardiovasc Imaging. 2011;4:210-219.) Key Words: myocardial infarction 䡲 MRI 䡲 edema 䡲 myocardial ischemia A purposes.8 –14 However, dark-blood T2-weighted MRI is prone to artifact, such as from motion or from blood stasis at the left ventricular wall leading to subendocardial bright rim artifacts. Imaging time in the cardiac cycle15 and coil sensitivity issues, which cause signal loss with depth of field,16 can give rise to diagnostic uncertainty with this method. Bright-blood T2-weighted MRI techniques have recently emerged as potential alternatives to dark-blood T2-weighted MRI. Kellman et al17 developed a bright-blood T2-prepared, single-shot steady-state free precession (T2-prepared SSFP) method, which involves surface coil intensity normalization, parallel imaging techniques, and motion-corrected averaging. fter acute myocardial infarction (MI), ex vivo1 and in vivo2 T2-weighted MRI enable estimation of the ischemic area at risk and myocardial salvage.3,4 T2-Weighted imaging improves the detection of acute MI5 and enables its discrimination from chronic MI.6 Overall, T2-weighted MRI method has emerging potential to guide the diagnosis and treatment of acute MI patients.7 Editorial see p 198 Clinical Perspective on p 219 Dark-blood inversion recovery T2-weighted MRI methods are well studied and widely used for clinical and research Received August 27, 2010; accepted March 15, 2011. From the West of Scotland Heart and Lung Centre, Golden Jubilee National Hospital, Glasgow, Scotland, United Kingdom (A.R.P., R.M., A.M., R.W., A.S., K.G.O., N.T., C.B.); BHF Glasgow Cardiovascular Research Centre, Institute of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow, Scotland, United Kingdom (A.R.P., M.C., J.M., N.T., C.B.); and Cardiovascular MR R&D, Siemens Healthcare, Erlangen, Germany, (X.B., S.Z.). Correspondence to Colin Berry, BSc, PhD, FRCP, FACC, BHF Glasgow Cardiovascular Research Centre, Institute of Cardiovascular and Medical Sciences 126 University Place, University of Glasgow, Glasgow, G12 8TA, Scotland, UK. E-mail [email protected] © 2011 American Heart Association, Inc. Circ Cardiovasc Imaging is available at http://circimaging.ahajournals.org 210 DOI: 10.1161/CIRCIMAGING.110.960450 Payne et al Bright-Blood T2-Weighted MRI Versus Dark-Blood MRI 211 Downloaded from http://circimaging.ahajournals.org/ by guest on May 11, 2017 Figure 1. A, MRI findings in a patient with acute STEMI. Matched diastolic cardiac MRI (left, cine MRI; middle left, phase-sensitive inversion recovery image; middle right, T2-weighted ACUT2E; right, dark-blood STIR) obtained in a 66-year-old man admitted with anterior STEMI. Pain to balloon time was 2 hours, 16 minutes. MRI was performed 15 hours after reperfusion. Anteroseptal transmural infarction (as revealed by late gadolinium enhancement, left middle) corresponds with transmural edema revealed by both bright-blood T2-weighted MRI (middle right) and dark-blood T2-weighted MRI (right). B, MRI findings in a 55-year-old female cigarette smoker 5 days after an acute NSTEMI. Coronary angiography 2 days earlier had revealed a culprit second obtuse marginal coronary artery that treated medically. Cine MRI (left) revealed a lateral left ventricular wall motion abnormality. Lateral nontransmural infarction (as revealed by late gadolinium enhancement, left middle) corresponds with transmural edema revealed by both bright-blood T2-weighted MRI (middle right) and dark-blood T2-weighted MRI (right). Aletras et al18 developed another bright-blood T2-weighted method, ACUT2E (Acquisition for Cardiac Unified T2 Edema), which is a hybrid of turbo spin-echo (TSE) and SSFP. The TSE-SSFP method has higher signal-to-noise (SNR) and contrast-to-noise (CNR) ratio than T2-prepared SSFP. We studied the diagnostic accuracy of bright-blood T2weighted MRI and dark-blood T2-weighted MRI method in acute MI by analyzing paired axial left ventricular slices. Our approach differed to that of Kellman et al,17 who used a delayed inversion recovery fast spin-echo method. We used the TSE-SSFP method because of its high SNR and CNR and compared this method with a standard dark-blood short tau inversion recovery (STIR) T2-weighted MRI method in 54 patients with acute MI. Methods Patient Population and Acute MI Management Fifty-four consecutive patients who underwent invasive treatment for acute MI at a regional cardiac center and who also had T2-weighted MRI were analyzed. No patients were excluded because of poor image quality. Exclusion criteria represented standard contraindications to contrast MRI, including an estimated glomerular filtration rate ⬍30 mL/min/1.73 m2. Our MRI research in acute MI has been approved by the West of Scotland Research Ethics Committee. The inclusion criteria included a diagnosis of acute MI based on a history of symptoms consistent with acute myocardial ischemia with supporting changes on the ECG, with or without regional STsegment elevation, associated with a typical rise in troponin I concentration.19 The culprit coronary artery was identified by coronary angiography. Acute MI management followed contemporary guidelines.19 Aspiration thrombectomy, direct stenting, antithrombotic drugs, and other therapies were administered according to clinical judgment. MRI Acquisition and Analyses MRI was performed on a Siemens MAGNETOM Avanto (Erlangen, Germany) 1.5-T scanner with an 8-element phased-array cardiac surface coil. The MRI protocol included SSFP cine MRI, T2weighted MRI with bright-blood ACUT2E TSE-SSFP,18 and darkblood STIR8 –12 and delayed enhancement phase-sensitive inversion recovery sequences.20 Axial left ventricular views were sequentially acquired with dark-blood and bright-blood T2-weighted MRI matched to the same slice position usually at the midpapillary level. Sample images are shown in Figure 1. The breath-hold bright-blood TSE-SSFP (ACUT2E) incorporates elements of the SSFP17 and TSE methods.8,17 This hybrid method, which does not involve a T2 preparation, has an SSFP readout gated during diastole and with a flip angle ␣ of 180°. There is a 90° preparation followed by 180° refocusing pulses, which is similar to TSE; however, this method is different from TSE because it involves gradient moment nulling rather than gradient crushing. TSE-SSFP creates a coherent train of T2-weighted spin-echos.18 Typical imaging parameters were acquisition time, 7 to 12 seconds; matrix, 192⫻192; flip angle, 180°; echo time (TE), 1.69 ms; and bandwidth, 789 Hz/pixel. Twenty-nine coherent spin-echoes (echo train length) were obtained per heartbeat, and the time interval (echo spacing) between the 180° inversion pulses was 3.4 ms. The trigger pulse was 2, such that data were acquired every second R-R interval. The voxel size was 1.9⫻1.9⫻6 mm3. 212 Circ Cardiovasc Imaging May 2011 Downloaded from http://circimaging.ahajournals.org/ by guest on May 11, 2017 The breath-hold black-blood T2 triple-inversion recovery pulse sequence (STIR, 8 to 12) involves a pair of slice-selective and nonselective 180° inversion pulses to null the blood pool signal followed by a third inversion pulse to null the fat signal. We used an 8-element phased-array surface coil. Surface coil intensity correction was routinely performed with prescan normalization and slicerelated shimming as appropriate to optimize dark-blood STIR acquisition. Postprocessing surface coil intensity correction was not performed. Typical imaging parameters were acquisition time, 8 to 13 seconds; matrix, 166⫻256; flip angle, 180°; effective echo time (TE), 61 ms; bandwidth, 235 Hz/pixel; turbo factor, 15; and trigger pulse, 2. The time interval between the 180° inversion pulses for STIR was 6.74 ms. The voxel size was 2.2⫻1.4⫻8 mm3. Microvascular obstruction (MVO) was defined as a dark zone on early delayed enhancement imaging 1, 3, 5, and 7 minutes after contrast injection and within an area of late gadolinium enhancement. MI was imaged using segmented phase-sensitive inversion recovery turbo fast low-angle shot,20 starting around 7 minutes after intravenous injection of 0.10 mmol/kg of gadoterate meglumine (Gd2⫹-DOTA, Dotarem, Guebert SA). Typical imaging parameters were matrix, 192⫻256; flip angle, 25°; TE, 3.36 ms; bandwidth, 130 Hz/pixel; echo spacing, 8.7 ms; and trigger pulse, 2. The voxel size was 1.8⫻1.3⫻8 mm3. MRI Analyses The images were analyzed on a Siemens workstation by MRI trained cardiologists with at least 5 years of experience. In addition, all 3 cardiologists had completed at least 2 years of full-time MRI research in patients with acute MI, including protocols with T2weighted MRI. A fourth cardiologist independently coordinated the images and data for the blinded analysis. Two cardiologists who did not take part in the blinded scoring for the presence or absence of edema performed quantitative assessments of infarct area and area at risk as determined by both T2-weighted methods. Left ventricular dimensions and volumes and ejection fraction were quantified using computer-assisted planimetry. Myocardial Edema and Standardized Measurements of T2-Weighted Area at Risk One bright-blood and 1 dark-blood T2-weighted MRI image acquired at the same slice position within the infarct zone were evaluated by each cardiologist. Hyperintense zones on T2-weighted MRI images were reviewed independently by 3 MRI-trained cardiologists who were blinded to the patients’ history and outcomes. Each observer scored for the presence or absence of edema and for the position and number of affected regions of the left ventricular slice for each patient. Segments were defined according to the 17-segment model of the American Heart Association.21 Disagreement was resolved by consensus. Quantitative assessments were performed by 2 independent observers, and their results were averaged. The jeopardized left ventricular area at risk on each axial image was defined as the percentage of left ventricular area delineated by the hyperintense zone on T2-weighted images.4 The window setting was defined as the sum of the mean myocardial signal intensity (SI) of the unaffected area plus 2 standard deviations (SD) for this area. The level setting was set at the mean SI of the unaffected area. This standardized approach was based on a recent clinical validation study.4 Myocardial tissue with a signal intensity with at least 2 SD above the mean signal obtained in the remote noninfarcted myocardium was considered the area at risk.4 Care was taken to exclude nonsuppressed blood pool signal because of slow flow adjacent to the subendocardium.3 Infarct Definition and Size The presence of acute infarction was established with MRI, based on abnormalities in cine wall motion, rest first-pass myocardial perfusion, and delayed-enhancement imaging. In addition, supporting changes on the ECG and coronary angiogram were also required. Acute infarction was considered present only if myocardial enhancement was confirmed on both the axial and long axis acquisitions. The myocardial mass of myocardial enhancement (grams) was quantified by a semiautomatic detection method using an SI threshold of ⬎2 SD above a remote reference region and expressed as a percentage of total left ventricular mass.22,23 MVO regions were included within the infarct area. Myocardial Salvage Myocardial salvage, as estimated by MRI, was calculated by subtraction of percent infarct size from percent area at risk.3,4 Biochemical Assessment of Infarct Size Troponin I was measured (AxSYM; Abbott) as a biochemical measure of infarct size. A blood sample was routinely obtained 12 to 24 hours after hospital admission. The troponin I protocol used in our hospital has an upper threshold of 22.8 g/L. Angiographic Jeopardy Scores The Lesion Score from the Alberta Provincial Project for Outcome Assessment in Coronary Heart Disease (APPROACH)24 was used to provide angiographic estimates of the area at risk.4 The angiograms were analyzed by a cardiologist who was independent of the MRI analyses. Statistical Analyses Normality was confirmed or excluded using the Shapiro-Francia test. Mean (SD) values and medians (interquartile range) were calculated. Correlations (r) between values for area at risk measured by MRI and the APPROACH Lesion Score were tested by Pearson or Spearman methods for normally and nonnormally distributed variables, respectively. All tests were 2-tailed. Comparisons of normally distributed continuous data between ST-segment–MI (STEMI) and non–STsegment–MI (NSTEMI) patient groups were undertaken using a Student t test. Between-group comparisons of nonnormally distributed data were performed with a Mann-Whitney test. Differences in the frequencies and proportions of clinical characteristics between patients with STEMI and NSTEMI were assessed with either a Fisher exact test or a 2 test with continuity correction, as appropriate. The McNemar exact test was used to compare observer agreement for the presence of regional myocardial edema with brightblood and dark-blood T2-weighted MRI methods. The level of agreement between area at risk and salvage estimated by brightblood and dark-blood T2-weighted MRI methods were assessed using Bland-Altman plots and 95% limits of agreement. The 95% limits of agreement are calculated using the mean difference between the 2 observers plus/minus twice the standard deviation of these differences and will contain approximately 95% of all such differences. Results Patient Characteristics Fifty-four patients with acute MI (44 (81%) STEMI) underwent MRI 2⫾2 days after initial treatment between October 1, 2009, and July 8, 2010 (Tables 1, 2, and 3). All patients had a history of acute ischemic symptoms, acute changes on the ECG, a typical rise in troponin I concentrations, a culprit coronary artery identified by coronary angiography, and typical findings on the contrast-enhanced MRI scan. MRI Findings The cardiac MRI findings are summarized in Tables 2, 3, 4, and 5. All patients had MRI evidence of segmental acute MI based on the presence of reduced wall motion and late gadolinium enhancement in the culprit artery territory. Five (9%) of these patients (4 STEMI, 1 NSTEMI) did not have Payne et al Table 1. Bright-Blood T2-Weighted MRI Versus Dark-Blood MRI Clinical Characteristics of Acute STEMI and NSTEMI Patients All (n⫽54) Mean⫾SD age, y Male sex, n (%) 58⫾10 38 (70) Acute STEMI (n⫽44, 81%) Acute NSTEMI (n⫽10, 19%) P 59⫾9 58⫾12 0.79 31 (70) 7 (70) 1.00 History, n (%) Previous MI 3 (6) 1 (2) 2 (20) 0.08 History of previous angina 7 (13) 2 (5) 2 (20) 0.15 History of PCI or CABG 0 (0) 0 (0) 0 (0) 1.00 History of chronic heart failure 1 (2) 0 (0) 1 (10) 0.18 38 (70) 33 (75) 5 (50) 0.14 2 (4) 1 (2) 1 (10) 0.34 1 46 (85) 38 (86) 8 (80) 2 5 (9) 3 (7) 2 (20) 3 3 (6) 3 (7) 0 (0) 4 0 (0) 0 (0) 0 (0) 54 (100%) 44 (100) 10 (100) ... 3.13 (2.37, 5.43) ... History of cigarette smoking Diabetes mellitus Presenting characteristics Heart failure, Killip class* Downloaded from http://circimaging.ahajournals.org/ by guest on May 11, 2017 Troponin-I elevation, n (%) ⬎upper limit of normal (threshold for positivity) (⬎0.04 g/L) 0.63 ... Cardiac catheter laboratory findings Median (IQR) time from onset of symptoms to primary PCI, h Infarct-related coronary artery Right 18 (33) 18 (41) 0 (0) Left anterior descending 28 (52) 20 (45) 8 (80) 8 (15) 6 (14) 2 (20) 0.045 27 (50) 22 (50) 5 (50) 1.00 ... 37 (10) ... ... 23 (43) 23 (52) 0 (0) Circumflex Multivessel coronary artery disease† APPROACH score25 TIMI flow grade at initial angiography 0 1 4 (7) 4 (9) 0 (0) 2 6 (11) 6 (14) 0 (0) 3 21 (39) 11 (25) 10 (100) 0.001 Treatment‡ PCI 46 (85) 43 (98) 3 (30) CABG 1 (2) 0 (0) 1 (10) Angiography and medical therapy 7 (13) 1 (2) 6 (60) ⬍0.001 Drug therapy during hospital admission, n (%) Aspirin 54 (100) 44 (100) 10 (100) 1.00 Glycoprotein IIbIIIa inhibitor 42 (78) 42 (95) 0 (0) ⬍0.001 Clopidogrel 48 (89) 43 (98) 5 (50) ⬍0.001 CABG indicates coronary artery bypass surgery; IQR, interquartile range; PCI, percutaneous coronary intervention; and TIMI, Thrombolysis In Myocardial Infarction. *Killip classes 2 through 4 were merged, and a Fisher exact test was used to compare the between group differences in proportions of patients in Killip class 1 versus Killip class 2 through 4. †Multivessel coronary artery disease was defined according to the presence of 2 or more arteries with stenoses of at least 50% of the reference vessel diameter, by visual assessment. ‡Treatment: CABG and PCI were merged and treated as 1 new variable (revascularization) and compared with medical therapy. 213 214 Circ Cardiovasc Imaging Table 2. May 2011 Cardiac MRI Findings All (n⫽54) Acute STEMI (n⫽44, 81%) Acute NSTEMI (n⫽10, 19%) P 50⫾12 51⫾11 47⫾15 0.44 Men 89⫾26 84⫾17 105⫾42 0.16 Women 76⫾15 76⫾17 75⫾3 0.72 Men 48⫾25 43⫾13 64⫾48 0.20 Women 35⫾15 35⫾17 35⫾1 1.0 54 (100) 44 (100) 10 (100) LV dimensions and function, mean⫾SD LV ejection fraction, % End-diastolic volume index, mL/m2 2 End-systolic volume index, mL/m Late gadolinium enhancement Patients with evidence of late gadolinium enhancement, n(%)* Downloaded from http://circimaging.ahajournals.org/ by guest on May 11, 2017 Patients categorized according to the No. of coronary artery territories with late enhancement in the slice position selected for T2-weighted MRI, n (%) None 5 (9) 4 (9) 1 (10) 43 (80) 36 (82) 7 (70) 6 (11) 4 (9) 2 (20) 42 (78) 36 (82) 6 (60) 0.20 Mean⫾SD acute infarct size, % 24.4⫾14.9 26.1⫾15.4 16.7⫾9.3 0.02 Microvascular obstruction, n (%) 42 (78) 35 (80) 7 (70) 0.67 Single coronary artery territory Multiple coronary artery territories Transmural late enhancement, n (%) 1.0 LV indicates left ventricular. *Late gadolinium enhancement within the axial slice coregistered to matched edema images; a coronary artery territory was assigned to the left anterior descending coronary artery, a diagonal branch, the circumflex coronary artery, an obtuse marginal artery branch or the right coronary artery. late enhancement on the slice selected for matched T2weighted imaging. MRI images from acute MI patients are shown in Figure 1 and 2. Detection of Acute MI by Bright-Blood and Dark-Blood T2-Weighted MRI Compared with dark-blood T2-weighted STIR, complete consensus agreement between 3 independent observers for recognition of myocardial edema in the infarct territory occurred in twice as many patients with bright-blood T2weighted TSE-SSFP (P⬍0.001; Table 3). When considered per segment of left ventricle, complete consensus was found in 85% of segments for bright blood but only 50% of segments for dark blood (P⬍0.001). Similar differences were found per patient and per segment when majority consensus was used. Bright-blood T2-weighted MRI had higher diagnostic accuracy for acute MI (Tables 4 and 5). Similarly, brightblood T2-weighted MRI had higher diagnostic accuracy (94%) than dark-blood T2-weighted MRI (61%) for identification of the culprit artery as revealed by coronary angiography (P⬍0.0001; Table 4). Dark-blood STIR had high sensitivity (96%) but weak specificity (35%) for detecting anterior MI (Table 5). Quantitative Assessment of Area at Risk and Myocardial Salvage In 19 patients (35%), the area at risk estimated using dark-blood STIR was less than infarct size (Figure 3). The average area at risk estimated by 2 independent observers with dark-blood T2-weighted MRI (28.7%⫾18.6%) was less than when measured by bright-blood T2-weighted MRI (37.9⫾13.6%; P⫽0.004). Consequently, myocardial salvage was also less when measured by dark-blood T2-weighted MRI (5.0⫾20.6%) compared with when measured by bright-blood T2-weighted MRI (13.8⫾11.2%; P⫽0.007). For area at risk estimated by 2 independent observers using bright-blood T2-weighted MRI, there was good correlation between observer 1 and observer 2 (r2⫽89%). The 95% limits of agreement were ⫺8.4% and 9.8% (Figure 4), and there was no evidence of bias (P⫽0.29). For area at risk estimated by STIR between these observers, there was poorer correlation (r2⫽59%), with 95% limits of agreement being ⫺32.8 and 26.2%. There was some evidence of bias, with observer 2 reporting a higher area at risk (P⫽0.05). The 95% limits of agreement for myocardial salvage (Figure 4) estimated by 2 independent observers using images obtained with bright-blood T2-weighted MRI were ⫺12% and 12%, respectively. There was no evidence of bias (P⫽0.96) on Bland-Altman analysis. The 95% limits of agreement for myocardial salvage estimated by these observers using images obtained with dark-blood T2weighted MRI were ⫺32% and 22%, respectively. There was evidence of bias (P⫽0.012) on Bland-Altman analysis and the confidence intervals for the bias between the 2 observers were ⫺8.7% to ⫺1.1%. Payne et al Bright-Blood T2-Weighted MRI Versus Dark-Blood MRI Table 3. Diagnostic Accuracy of Bright-Blood TSE-SSFP and Dark-Blood T2-Weighted STIR for Detection of Myocardial Edema in the Infarct Territory* as Independently Assessed by 3 Observers per Patient and Per Segment† of Left Ventricle Table 4. Identification of the Culprit Coronary Artery Based on the Presence of Regional Hyperenhancement on T2-Weighted MRI According to the Consensus Formed by 3 Independent Observers Dark Blood STIR MRI Consensus agreement per patient (n⫽49 with segmental delayed enhancement)* Full consensus Complete agreement of all 3 observers Bright-blood T2-weighted MRI Dark-Blood T2-Weighted MRI Correct Incorrect P ⬍0.001 215 Culprit Artery Territory Identified Correct (n⫽33) Incorrect (n⫽21) Correct (n⫽51) 33 18 Incorrect (n⫽3) 51 (94%) 0 3 3 (6%) 33 (61%) 21 (39%) P⬍0.0001 Bright-blood T2-weighted MRI Correct 16 20 Incorrect 2 11 Majority consensus blood STIR (33⫾18%) and ACUT2E (39⫾18%) were similar (P⫽0.2). Correlation of Area at Risk Estimated by T2-Weighted MRI Versus an Angiographic Score Agreement with ⱖ2 observers Dark-Blood T2-Weighted MRI Downloaded from http://circimaging.ahajournals.org/ by guest on May 11, 2017 Correct Incorrect Correct 25 21 Incorrect 1 2 Bright-blood T2-weighted MRI ⬍0.001 Consensus agreement per segment of left ventricle (n⫽96 segments with acute MI)† Full consensus Complete agreement of all 3 observers Dark-Blood T2-Weighted MRI Correct Incorrect P Bright-blood T2-weighted MRI Correct 44 38 Incorrect 4 10 ⬍0.001 Majority consensus Agreement with ⱖ2 observers Dark-Blood T2-Weighted MRI Correct Incorrect Correct 60 33 Incorrect 3 0 Bright-blood T2-weighted MRI ⬍0.001 *The presence and location of acute infarction was established based on the presence of reduced regional wall motion and a perfusion abnormality on first pass associated with late gadolinium enhancement, also supported by acute changes on the ECG and the culprit artery on coronary angiography. Based on these criteria, there were 96 segments with evidence of acute MI including evidence of late gadolinium enhancement in 49 patients. †The left ventricle was segmented according to the American Heart Association model.21 Comparison of Area at Risk Estimated by Bright-Blood T2-Weighted MRI Versus Dark-Blood STIR In patients who underwent MRI ⬍2 days after MI (n⫽38), the area at risk estimated by dark-blood STIR was less than the area at risk estimated by ACUT2E (27⫾19% versus 37⫾12%; P⫽0.002). In patients who underwent MRI ⱖ2 days after MI (n⫽12), the areas at risk estimated by dark- The area at risk estimated by the APPROACH Lesion Score (Table 1) was 37⫾10%. The area at risk estimated by brightblood ACUT2E correlated reasonably well with area at risk estimated by the APPROACH Lesion Score (r⫽0.62, P⬍0.01). The correlations between the area at risk estimated using a single matched axial image from the ACUT2E and STIR methods versus the APPROACH Lesion Score and were r⫽0.68 and 0.44, both P⬍0.01. Bright rim artifacts occurred in 9 (17%) patients with dark-blood T2-weighted MRI; however, these artifacts were not associated with false-positive or false-negative results. Discussion The main findings of this study are that bright-blood T2weighted MRI has higher diagnostic accuracy for detection of acute MI than dark-blood STIR, which also underestimates the ischemic area at risk and myocardial salvage. Additionally, interobserver agreement was better with bright-blood T2-weighted MRI. Since T2-weighted MRI became available for use clinically,8 dark-blood2,5–14 techniques have become established for the detection and assessment of myocardial edema. However, diagnostic uncertainty with dark-blood T2weighted STIR (related mainly to surface coil sensitivity problems, imaging timing and motion artifact15) has led to alternative approaches. One development with dark-blood STIR has been the use of the body coil instead of a chest wall surface coil,2,13,14 at the expense of reduced SNR. Use of a surface coil normalization algorithm may also help. More recently, new bright-blood T2-weighted techniques with SSFP have been developed,4,17,18 and their emergence has stimulated us to question the comparative diagnostic accuracy of bright-blood versus dark-blood T2-weighted MRI. Our results extend those of Kellman et al,17 who found that bright-blood T2-prepared SSFP had higher diagnostic accuracy than dark-blood TSE in 22 acute MI patients. We aimed to provide further information in a larger patient cohort by comparing another bright-blood T2-weighted method, ACUT2E TSE-SSFP,18 with a standard black-blood TSE method. The bright-blood ACUT2E TSE-SSFP method has higher SNR and CNR compared with T2-prepared SSFP.17 Our study included a larger number of patients with acute 216 Circ Cardiovasc Imaging May 2011 Table 5. Sensitivity and Specificity of Bright-Blood TSE-SSFP and Dark-Blood STIR MRI for Detection of Anterior Versus Nonanterior MI Bright Blood T2-Weighted MRI Anterior MI (n⫽28) Nonanterior MI (n⫽26) Anterior Edema Nonanterior Edema 27 1 0 26 Dark Blood T2-Weighted MRI Anterior Edema Nonanterior Edema Sensitivity⫽96% 27 1 Sensitivity⫽96% Specificity⫽100% 17 9 Specificity⫽35% Downloaded from http://circimaging.ahajournals.org/ by guest on May 11, 2017 coronary syndromes thus extending the observations by Kellman et al17 to those with non–ST elevation MI. The dark-blood preparation used in TSE T2-weighted imaging introduces potential signal loss due reduced surface coil sensitivity with depth of field and also through-plane cardiac motion.15–16 Our results indicate that this problem mainly affects detection of acute myocardial injury in the basal and mid inferior as well as posterior left ventricular walls. Indeed, dark-blood T2 STIR had reduced diagnostic accuracy for nonanterior MI compared with anterior MI. Another intriguing finding was that in nearly one-third of acute MI patients, the area at risk estimated by dark-blood T2-weighted was less than infarct size, suggesting that the dark-blood T2-weighted may underestimate area at risk in some patients. Interobserver agreement was better with bright-blood T2-weighted MRI than dark-blood STIR. Because quantitative assessment of the ischemic area at risk and myocardial salvage have emerging importance for clinical and research purposes,3–15 our results call into question the use of dark-blood STIR even with surface coil normalization in acute MI patients. Optimized timing of the TSE read-out can improve problems related to cardiac motion15; however, this is harder to optimize at higher heart rates, especially with variation in Figure 2. A, Dark-blood STIR (right) reveals a hyperintense area in the anterolateral left ventricular wall in a patient with inferolateral MI. MRI findings in a patient with acute inferolateral STEMI. A 60-year-old male smoker presented with inferolateral STEMI. Primary percutaneous coronary intervention (PCI) was performed in an occluded dominant left circumflex coronary artery. The pain to balloon time was 2 hours, 17 minutes. Coronary artery blood flow was mildly reduced (TIMI flow grade 2) at end of the PCI. Cardiac MRI was performed 19 hours after hospital admission. Inferolateral transmural infarction with MVO (as revealed by late gadolinium enhancement, left middle) corresponds with transmural edema revealed by bright-blood T2-weighted MRI (middle right). A hyperintense zone in the anterolateral wall on dark-blood T2-weighted MRI (right) incorrectly suggests anterior MI and involvement of the left anterior descending coronary artery. B, Inferior MI that is not apparent with dark-blood STIR (right). A 52-year-old obese male smoker presented with inferior STEMI. Coronary angiography revealed a distal occlusion of his right coronary artery. The pain to balloon time was 83 minutes. MRI was performed 14 hours after hospital admission. Inferior nontransmural infarction with MVO (as revealed by late gadolinium enhancement, left middle) corresponds with transmural edema revealed by bright-blood T2-weighted MRI (middle right) but not by dark-blood T2-weighted MRI (right). C, Subendocardial bright rim artifact with dark-blood STIR. A 48-year-old male smoker presented with an anterior STEMI and underwent primary PCI in the culprit left anterior descending coronary artery. MRI was performed 14 hours after hospital admission and revealed anteroseptal transmural infarction with MVO (as revealed by late gadolinium enhancement, left middle) corresponds with transmural edema revealed by both bright-blood T2-weighted MRI (middle right) and dark-blood T2-weighted MRI (right). An anterior subendocardial bright rim artifact is present in the dark-blood STIR image. Payne et al Bright-Blood T2-Weighted MRI Versus Dark-Blood MRI 217 For the edema image analyses to enable estimation of the area at risk, we adopted a standardized approach for window and level intensity settings that had been previously worked out in an earlier clinical validation study.4 Standardized window and level settings were based on the signal intensity of normal (remote myocardium) and a signal intensity difference of 2 SD was used to discriminate affected (hyperintense) myocardium from unaffected zones. Single-shot T2-prepared SSFP can be acquired during free breathing and multiple images may be motion-corrected and averaged to enhance the SNR. This method provides a uniform proton density field map and can be used for T2 mapping to measure absolute values of T2 in the heart. T2-prepared SSFP is an attractive alternative method for bright-blood edema imaging.17 In the future, quantitative T2 mapping could help to improve detection of regional myocardial edema compared with visual assessment alone.25 Downloaded from http://circimaging.ahajournals.org/ by guest on May 11, 2017 Limitations Figure 3. Scatterplots with lines of identity for the relationships between infarct size measured by MRI versus area at risk derived with A, bright-blood T2-weighted MRI, and B, darkblood T2-weighted MRI. Average values of area at risk were obtained from measurements from 2 independent observers. Because infarct size is physiologically a subset of the area at risk, the expected relationship between area at risk and infarct size involves all points at or below the line of identity. This is so for all cases with bright-blood T2-weighted MRI but only 35 (65%) of cases with dark-blood T2-weighted MRI (P⬍0.001, Fisher exact test). R-R intervals. Furthermore, timing optimization complicates the examination. Motion tracking can further improve image quality.16 Subendocardial rim artifacts have the potential to be problematic in acute MI patients because of reduced regional wall motion, particularly at the left ventricular apex. We found that bright rim artifacts were fairly common. We also found that the presence of these artifacts was not associated with diagnostic accuracy. This could be because the observers were able to correctly discriminate bright rim artifacts from edema. It is also possible that meticulous slice-related shimming during dark-blood STIR acquisition reduced their intensity. We aimed to use the T2-weighted MRI methods in a way that closely reflected standard (usual) clinical practice. Therefore, we used the methods as set-up to be used on our scanner. Furthermore, there are other differences inherent to these methods (eg, different spatial resolutions) that we chose to accept rather than modify. For example, shimming was not routinely used with the ACUT2E method, whereas shimming is standard practice for image optimization with STIR. Also, breath-hold duration is shorter with the ACUT2E method than with STIR, and, because long breath-holds are associated with reduced image quality, this could be one other source of differences in image quality. Again, we did not attempt to alter breath-hold durations. Rather, we accepted the methods as set on the scanner. Because of time constraints when imaging our acute MI patients, we focused on 2 T2-weighted MRI methods. All patients had evidence of late gadolinium enhancement, but 5 of these patients did not have late enhancement in the image position that had been selected for T2-weighted MRI. This can be explained by the fact that T2-weighted MRI was obtained in the culprit area of injured myocardium revealed by reduced wall motion but before gadolinium contrast administration. Therefore, the T2 MRI acquisitions in these 5 patients can be interpreted as having been acquired in a territory of noninfarcted but injured myocardium evidenced by wall motion and edema. Coil sensitivity corrections with the ACUT2E method may be imperfect because of nonuniform proton density field map. Furthermore, neither ACUT2E nor STIR can be used for T2 mapping and, like STIR, ACUT2E is dependent on subjective image interpretation. T2-Dark-blood STIR had lower sensitivity in accurately identifying the area at risk in patients with acute coronary syndrome. Whether the addition of coil profile normalization and/or T2 mapping could have improved its accuracy has not been addressed in this study. We suggest that the addition of complicated algorithms to resolve a potential weakness of a test complicates rather than facilitates the clinical applicability and appeal of the diagnostic method. For several reasons, it has not been possible to estimate CNR and SNR in our analysis. The STIR method does not compensate for coil sensitivity variation across the imaging plane and because TSE methods are sensitive to motion, T2 signal differences may arise between injured (reduced motion) and uninjured (normal motion) myocardium in addition to contributions from T2 differences related to myocardial edema. Noise measurements are not incorporated in the ACUT2E method, and retrospective measurements are not possible with parallel imaging because noise values may be variable across the field of view. Because all of the patients in our cohort had a confirmed diagnosis of acute MI, the study design only permits an evaluation of the sensitivity and not the specificity of the 218 Circ Cardiovasc Imaging May 2011 Downloaded from http://circimaging.ahajournals.org/ by guest on May 11, 2017 Figure 4. Bland-Altman plots for A, area at risk, and B, salvage in 54 acute MI subjects revealed by bright-blood T2-weighted MRI (upper row) and darkblood T2-weighted MRI (lower row) measured by 2 independent observers. The 95% limits of agreement for area at risk estimation using images obtained with bright-blood T2-weighted MRI were ⫺8.5% and 9.9%, respectively. There was no evidence of bias (P⫽0.29) on Bland-Altman analysis. The 95% limits of agreement for area at risk estimation using images obtained with dark-blood T2-weighted MRI were ⫺33% and 26%, respectively. There was no evidence of bias (P⫽0.113) on Bland-Altman analysis. The 95% limits of agreement for myocardial salvage estimation using images obtained with bright-blood T2-weighted MRI were ⫺12% and 12%, respectively. There was no evidence of bias (P⫽0.9) on Bland-Altman analysis. The 95% limits of agreement for myocardial salvage estimation using images obtained with dark-blood T2-weighted MRI were ⫺33% and 23%, respectively. There was evidence of bias (P⫽0.012) on Bland-Altman analysis. methods. Our results demonstrate that bright-blood MRI increases the rate of detection of acute MI cases therefore raising sensitivity. Although our results do not provide definitive information on whether or not bright-blood T2weighted MRI might raise the rate of falsely reporting edema in patients without acute MI (ie, reducing specificity), the fact that uninjured segments from myocardium remote from the injured culprit artery territory were also included in our analysis provides some reassurance that this may not be the case. Anticipated Value Our findings indicate that the diagnostic utility of dark-blood STIR has some limitations when used in patients with known or suspected acute MI. Based on our findings, bright-blood T2-weighted MRI appears to have acceptable diagnostic utility. Acknowledgments We thank the staff in the Imaging Centre in the Golden Jubilee National Hospital, and in particular Jennifer Gilchrist and Carolyn Clark. We also thank the staff in the Cardiac Catheterization Laboratories and Coronary Care Unit in our hospital. We acknowledge the excellent technical support from Ms Christie McComb and Dr John Foster, Department of Medical Physics, Greater Glasgow and Clyde Health Board. We gratefully acknowledge the support of Dr Patrick Revell and Dr Craig Buckley from MR R&D, Siemens Healthcare, UK. Sources of Funding Dr Berry holds a Senior Fellowship from the Scottish Funding Council. This research was supported by a grant from the Chief Scientist Office, Scottish Health Department. Disclosures Drs Bi and Zuehlsdorff are employees of Siemens Healthcare, the manufacturer of the MRI scanner and pulse sequences used in this study. References 1. Garcia-Dorado D, Oliveras J, Gili J, Sanz E, Perezvilla F, Barrabes J, Carreras MJ, Solares J, Soler-Soler J. Analysis of myocardial edema by magnetic-resonance-imaging early after coronary-artery occlusion with or without reperfusion. Cardiovasc Res. 1993;27:1462–1469. 2. Aletras AH, Tilak GS, Natanzon A, Hsu LY, Gonzalez FM, Hoyt RF Jr, Arai AE. Retrospective determination of the area at risk for reperfused acute myocardial infarction with T2-weighted cardiac magnetic resonance imaging: histopathological and displacement encoding with stimulated echoes (DENSE) functional validations. Circulation. 2006 18;113: 1865–1870. 3. Friedrich MG, Abdel-Aty H, Taylor A, Schulz-Menger J, Messroghli D, Dietz R. The salvaged area at risk in reperfused acute myocardial infarction as visualized by cardiovascular magnetic resonance. J Am Coll Cardiol. 2008;51:1581–1587. 4. Berry C, Kellman P, Mancini C, Chen MY, Bandettini WP, Lowrey T, Hsu LY, Aletras AH, Arai AE. Magnetic resonance imaging delineates the ischemic area-at-risk and myocardial salvage in patients with acute myocardial infarction. Circ Cardiovasc Imaging. 2010;3:527–535. 5. Cury RC, Shash K, Nagurney JT, Rosito G, Shapiro MD, Nomura CH, Abbara S, Bamberg F, Ferencik M, Schmidt EJ, Brown DF, Hoffmann U, Brady TJ. Cardiac magnetic resonance with T2-weighted imaging improves detection of patients with acute coronary syndrome in the emergency department. Circulation. 2008;118:837– 844. 6. Abdel-Aty H, Zagrosek A, Schulz-Menger J, Taylor AJ, Messroghli D, Kumar A, Gross M, Dietz R, Friedrich MG. Delayed enhancement and T2-weighted cardiovascular magnetic resonance imaging differentiate acute from chronic myocardial infarction. Circulation. 2004;109: 2411–2416. 7. Friedrich MG. Myocardial edema: a new clinical entity? Nat Rev Cardiol. 2010;7:292–296. Payne et al Bright-Blood T2-Weighted MRI Versus Dark-Blood MRI Downloaded from http://circimaging.ahajournals.org/ by guest on May 11, 2017 8. Simonetti OP, Finn JP, White RD, Laub G, Henry DA. “Black blood” T2-weighted inversion-recovery MR imaging of the heart. Radiology. 1996;199:49 –57. 9. Francone M, Bucciarelli-Ducci C, Carbone I, Canali E, Scardala R, Calabrese FA, Sardella G, Mancone M, Catalano C, Fedele F, Passariello R, Bogaert J, Agati L. Impact of primary coronary angioplasty delay on myocardial salvage, infarct size, and microvascular damage in patients with ST-segment elevation myocardial infarction: insight from cardiovascular magnetic resonance. J Am Coll Cardiol. 2009;54:2145–2153. 10. O’Regan DP, Ahmed R, Neuwirth C, Tan Y, Durighel G, Hajnal JV, Nadra I, Corbett SJ, Cook SA. Cardiac MRI of myocardial salvage at the peri-infarct border zones after primary coronary intervention. Am J Physiol Heart Circ Physiol. 2009;297:H340 –H346. 11. Ganame J, Messalli G, Dymarkowski S, Rademakers FE, Desmet W, Van de Werf F, Bogaert J. Impact of myocardial haemorrhage on left ventricular function and remodelling in patients with reperfused acute myocardial infarction. Eur Heart J. 2009;30:1440 –1449. 12. Raman SV, Simonetti OP, Winner MW III, Dickerson JA, He X, Mazzaferri EL Jr, Ambrosio G. Cardiac magnetic resonance with edema imaging identifies myocardium at risk and predicts worse outcome in patients with non-ST-segment elevation acute coronary syndrome. J Am Coll Cardiol. 2010;55:2480 –2488. 13. Larose E, Rodés-Cabau J, Pibarot P, Rinfret S, Proulx G, Nguyen CM, Déry JP, Gleeton O, Roy L, Noël B, Barbeau G, Rouleau J, Boudreault JR, Amyot M, De Larochellière R, Bertrand OF. Predicting late myocardial recovery and outcomes in the early hours of ST-segment elevation myocardial infarction traditional measures compared with microvascular obstruction, salvaged myocardium, and necrosis characteristics by cardiovascular magnetic resonance. J Am Coll Cardiol. 2010;55:2459 –2469. 14. Eitel I, Desch S, Fuernau G, Hildebrand L, Gutberlet M, Schuler G, Thiele H. Prognostic significance and determinants of myocardial salvage assessed by cardiovascular magnetic resonance in acute reperfused myocardial infarction. J Am Coll Cardiol. 2010;55:2470 –2479. 15. Wince WB, Kim RJ. Molecular imaging: T2-weighted CMR of the area at risk: a risky business? Nature Rev Cardiol. 2010;7:547–549. 16. Keegan J, Gatehouse PD, Prasad SK, Firmin DN. Improved turbo spin echo imaging of the heart with motion-tracking. J Magn Reson Imaging. 2006;24:563–570. 219 17. Kellman P, Aletras AH, Mancini C, McVeigh ER, Arai AE. T2-prepared SSFP improves diagnostic confidence in edema imaging in acute myocardial infarction compared to turbo spin echo. Magn Reson Med. 2007; 57:891– 897. 18. Aletras AH, Kellman P, Derbyshire JA, Arai AE. ACUT2E TSE-SSFP: a hybrid method for T2-weighted imaging of edema in the heart. Magn Reson Med. 2008;59:229 –235. 19. Antman EM, Anbe DT, Armstrong PW, Bates ER, Green LA, Hand M, Hochman JS, Krumholz HM, Krushner FG, Lamas GA. ACC-AHA guidelines for the management of patients with ST-elevation myocardial infarction. J Am Coll Cardiol. 2004;44:E1–E211. 20. Kellman P, Arai AE, McVeigh ER, Aletras AH. Phase-sensitive inversion recovery for detecting myocardial infarction using gadolinium-delayed hyperenhancement. Magn Reson Med. 2002;47:372–383. 21. Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, Pennell DJ, Rumberger JA, Ryan T, Verani MS. 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. 2002;105:539 –542. 22. McGeoch R, Watkins S, Steedman T, Berry C, Davie A, Byrne J, Hillis WS, Lindsay MM, Robb SD, Dargie HJ, Oldroyd KG. The index of microvascular resistance measured acutely predicts the severity and nature of myocardial Infarction in patients with ST segment elevation myocardial infarction. J Am Coll Cardiol Cardiovasc Interv. 2010;3: 715–722. 23. Weir RA, Murphy CA, Petrie CJ, Martin TN, Balmain S, Clements S, Steedman T, Wagner GS, Dargie HJ, McMurray JJV. Microvascular obstruction remains a portent of adverse remodeling in optimally-treated patients with left ventricular systolic dysfunction after acute myocardial infarction. Circ Imaging. 2010;3:360 –367. 24. Graham MM, Faris PD, Ghali WA, Galbraith PD, Norris CM, Badry JT, Mitchell LB, Curtis MJ, Knudtson ML, APPROACH Investigators (Alberta Provincial Project for Outcome Assessment in Coronary Heart Disease). Validation of three myocardial jeopardy scores in a population based cardiac catheterization cohort. Am Heart J. 2001;142:254 –261. 25. Giri S, Chung YC, Merchant A, Mihai G, Rajagopalan S, Raman SV, Simonetti OP. T2 quantification for improved detection of myocardial edema. J Cardiovasc Magn Reson. 2009;11:56. CLINICAL PERSPECTIVE T2-weighted MRI has diagnostic and clinical utility for detecting myocardial edema in patients with known or suspected acute myocardial infarction. T2-weighted MRI also enables estimation of the ischemic area at risk and myocardial salvage in patients with acute myocardial infarction. The T2-weighted MRI method that is most widely used is dark-blood short tau inversion recovery (STIR). However, dark-blood STIR is prone to artifact from motion, subendocardial blood stasis, and technical issues with scan timing or surface coil sensitivity. All of these problems with dark-blood STIR may contribute to diagnostic uncertainty with this method. Based on recent technical developments, an alternative bright-blood T2-weighted MRI approach has emerged. Therefore, we compared the diagnostic accuracy of a new bright-blood T2-weighted MRI method with standard dark-blood STIR in patients with acute myocardial infarction. We found that compared with dark-blood STIR, bright-blood T2-weighted MRI was associated with higher consensus agreements between observers for identification of myocardial edema and a higher rate of detection of the culprit coronary artery. The dark-blood method also underestimated the area at risk compared with infarct size. Our results support the use of bright-blood T2-weighted MRI instead of STIR for the assessment of myocardial edema. Downloaded from http://circimaging.ahajournals.org/ by guest on May 11, 2017 Bright-Blood T2-Weighted MRI Has Higher Diagnostic Accuracy Than Dark-Blood Short Tau Inversion Recovery MRI for Detection of Acute Myocardial Infarction and for Assessment of the Ischemic Area at Risk and Myocardial Salvage Alexander R. Payne, Matthew Casey, John McClure, Ross McGeoch, Aengus Murphy, Rosemary Woodward, Andrew Saul, Xiaoming Bi, Sven Zuehlsdorff, Keith G. Oldroyd, Niko Tzemos and Colin Berry Circ Cardiovasc Imaging. 2011;4:210-219; originally published online March 22, 2011; doi: 10.1161/CIRCIMAGING.110.960450 Circulation: Cardiovascular Imaging is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright © 2011 American Heart Association, Inc. All rights reserved. Print ISSN: 1941-9651. Online ISSN: 1942-0080 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://circimaging.ahajournals.org/content/4/3/210 Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Circulation: Cardiovascular Imaging can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Circulation: Cardiovascular Imaging is online at: http://circimaging.ahajournals.org//subscriptions/