Master the cardiac MRI protocol for function and morphology: ECG-gated cine SSFP, gadolinium dosing for LGE, arrhythmia-gating fixes, and ejection-fraction pitfalls.

Cardiac MRI Protocol: Mastering Function and Morphology Imaging

Introduction

The cardiac MRI protocol for function and morphology is the diagnostic backbone of modern cardiovascular imaging, combining exquisite soft-tissue contrast with true volumetric accuracy in a single, radiation-free examination. Unlike echocardiography, cardiac magnetic resonance (CMR) is not limited by acoustic windows, and unlike coronary CT angiography, it requires no ionizing radiation to characterize both chamber function and myocardial tissue composition. For radiographers, radiologists, and hospital administrators alike, understanding this protocol is essential: it underpins the diagnostic pathway for cardiomyopathy, ischemic heart disease, myocarditis, and infiltrative disorders that echocardiography alone frequently cannot resolve.

This tenth instalment of the 30-Day MRI Protocol Mastery Series focuses on Cardiac (Function & Morphology) imaging — the workhorse protocol built around ECG-gated segmented cine balanced steady-state free precession (bSSFP), regional shimming over the left ventricle, and phase-sensitive inversion-recovery (PSIR) late gadolinium enhancement, following standardized Society for Cardiovascular Magnetic Resonance protocols.^[1] The single greatest technical threat to this protocol is ECG misgating in the setting of arrhythmia, and much of this article is dedicated to understanding, preventing, and correcting that failure mode.

For hospital administrators, the operational stakes are equally significant: a cardiac MRI slot typically runs 45–60 minutes, considerably longer than most other MRI examinations, and a repeat scan triggered by unrecognized misgating or poor contrast timing represents a substantial throughput and cost burden. Investing in staff competency, consistent contrast-delivery infrastructure, and arrhythmia-aware acquisition options is therefore as much a capacity-planning decision as a clinical-quality one.

Anatomy of the heart for CMR

The heart is a four-chambered muscular pump oriented obliquely within the mediastinum, with its long axis running from the right hip toward the left shoulder. This oblique orientation means that true cardiac imaging planes never align with the conventional axial, sagittal, or coronal body planes; instead, every cardiac MRI protocol is built from a cascade of double-oblique localizers that progressively isolate the vertical long axis (2-chamber), horizontal long axis (4-chamber), and short-axis stack.

The left ventricle

The left ventricle (LV) is the primary functional target of this protocol. Its wall is conventionally divided into the 17-segment American Heart Association model, spanning basal, mid-cavity, and apical short-axis levels plus the true apex. Segmental analysis of wall thickness, wall motion, and late gadolinium enhancement distribution against this 17-segment map is fundamental to correlating imaging findings with coronary artery territories (left anterior descending, left circumflex, right coronary artery).^[2] Population-derived normal reference ranges for chamber volumes and mass are essential for correct interpretation.^[21]

The right ventricle

The right ventricle (RV) has a complex, crescentic geometry that makes short-axis-only volumetry unreliable; dedicated RV-focused stacks or axial cine acquisitions are frequently added when arrhythmogenic right ventricular cardiomyopathy (ARVC) or pulmonary hypertension is suspected. The RV free wall is thin and trabeculated, complicating both endocardial border delineation and the detection of subtle regional wall-motion abnormalities.

Atria, valves, and pericardium

The atria are assessed primarily for volume and for the presence of thrombus (particularly the left atrial appendage), while the four cardiac valves are evaluated qualitatively on cine imaging for regurgitant or stenotic jets. Cine bSSFP readily depicts a signal void jet across an incompetent or stenotic valve, and while CMR is not the primary modality for valvular disease, incidental valvular findings are common and should be described using consistent terminology (mild/moderate/severe) rather than left unremarked. The pericardium, a thin fibroserous sac normally under 2 mm in thickness, is best assessed on black-blood or cine sequences, and its thickening or effusion has direct implications for the differentiation of constrictive pericarditis from restrictive cardiomyopathy.^[24]

Coronary artery territories and the aortic root

Because LGE distribution is interpreted against the 17-segment model, radiographers and radiologists both benefit from an explicit map of expected coronary territories: the left anterior descending (LAD) artery typically supplies the anteroseptal and anterior segments, the left circumflex (LCx) supplies the inferolateral and lateral segments, and the right coronary artery (RCA) supplies the inferior and inferoseptal segments plus, in a right-dominant system, the posterior wall. The aortic root and proximal ascending aorta are included on long-axis localizers primarily to exclude root dilation or dissection as an alternative cause of chest pain.

The myocardium as a tissue

Beyond gross anatomy, the myocardium itself is a compound tissue of cardiomyocytes, interstitial collagen, and microvasculature. Diffuse or focal expansion of the interstitial space — from fibrosis, edema, amyloid infiltration, or iron deposition — alters the native T1, T2, and T2* relaxation properties of the tissue long before any change is visible on cine imaging, which is precisely why quantitative mapping sequences have become integral to a comprehensive cardiac MRI protocol.^[3][7]

MR tissue relaxation values

Quantitative mapping has shifted cardiac MRI from a purely qualitative, pattern-recognition discipline toward one with reproducible, scanner-independent reference values. The table below summarizes the accepted normal-range relaxation values for the principal cardiac and mediastinal tissues at 1.5T and 3.0T.^[3]

These values are acquired using modified Look-Locker inversion recovery (MOLLI) or shortened MOLLI (ShMOLLI) sequences for T1, and T2-prepared bSSFP for T2. Extracellular volume fraction (ECV), derived from pre- and post-contrast T1 combined with hematocrit, provides an additional quantitative biomarker of diffuse fibrosis and amyloid burden that is independent of scanner-specific T1 offsets.^[16][17]

Scanning technique

A complete cardiac MRI protocol for function and morphology proceeds through a fixed sequence of steps, each dependent on accurate ECG signal quality from the outset.^[1]

1. Patient preparation and ECG lead placement. Skin is exfoliated and vector-ECG electrodes are placed in a diamond configuration to maximize R-wave amplitude and minimize magnetohydrodynamic T-wave distortion.
2. Coil setup and isocentering. A dedicated cardiac phased-array coil is centered over the heart; the patient is positioned supine with arms above the head to reduce wraparound artifact.
3. Survey and localizer acquisition. Axial, sagittal, and coronal localizers establish the initial cardiac position.
4. Vertical long-axis (2-chamber) plane. Prescribed from the axial localizer through the LV apex and mitral valve.
5. Horizontal long-axis (4-chamber) plane. Prescribed from the 2-chamber view, bisecting both ventricles and atria.
6. Short-axis stack planning. Contiguous 6–8 mm slices are prescribed perpendicular to the LV long axis, covering base to apex.
7. Cine bSSFP acquisition. Segmented, ECG-gated cine images are acquired in each plane, typically during 8–12 second breath-holds, with local shimming applied over the LV to reduce off-resonance banding artifact.
8. Tissue characterization sequences. T1 mapping (pre-contrast), T2 mapping or STIR, and T2* (if iron overload is suspected) are acquired prior to contrast administration.
9. Contrast administration and post-contrast T1 mapping. Gadolinium is injected per the protocol below; post-contrast T1 mapping enables ECV calculation.
10. Late gadolinium enhancement (LGE) acquisition. A Look-Locker TI scout identifies the myocardial null point, followed by PSIR LGE imaging in short- and long-axis planes 10–15 minutes post-injection.

In practice, steps 4–6 (long-axis and short-axis plane prescription) are the most technically demanding for less-experienced radiographers, since each subsequent plane is prescribed from the previous one rather than from a fixed anatomical landmark. An error of even 10–15 degrees in the initial 2-chamber plane propagates into an off-axis short-axis stack, which in turn produces systematically inaccurate ejection-fraction and mass calculations. Departmental competency programs should therefore prioritize supervised repetition of this planning cascade before independent cardiac scanning is authorized.

Scanner comparison: 1.5T versus 3.0T

Contrast media protocol

Gadolinium-based contrast is administered for two complementary purposes in this cardiac MRI protocol: first-pass perfusion assessment (where clinically indicated) and, universally, late gadolinium enhancement for scar and fibrosis detection.^[22]

Gadolinium-based contrast agents are classified by the ACR into Group I (linear, associated with the highest reported nephrogenic systemic fibrosis risk), Group II (macrocyclic and protein-stable linear agents, with the lowest reported risk and preferred for routine cardiac use), and Group III (agents with limited unconfounded NSF data). Departmental formularies should default to a single Group II agent for cardiac protocols to simplify dosing calculations, TI-nulling behavior, and staff familiarity across the full 20–30 mL dose range used in this protocol.

Because myocardial nulling changes progressively as contrast washes out of the extracellular space, the TI must be re-optimized if there is a delay between the scout and the diagnostic LGE acquisition. Using a stale TI value is one of the most common technical causes of false-negative or false-positive scar assessment.

Radiofrequency power deposition (SAR)

Unlike CT protocols, this cardiac MRI protocol carries no ionizing radiation dose; the relevant physical safety constraint is instead the Specific Absorption Rate (SAR) — the rate of radiofrequency energy deposition in tissue, measured in W/kg. This is governed internationally by IEC 60601-2-33, ICNIRP exposure guidelines, and FDA guidance documents rather than the ionizing-radiation frameworks (ICRP, EC RP 185, AAPM) that apply to CT protocols elsewhere in this series; those frameworks are referenced here only insofar as they inform the broader radiological safety culture common to both modalities.^[13][14]

Five SAR-reduction strategies for cardiac MRI

1. Reduce flip angle in bSSFP cine sequences, particularly at 3.0T, while preserving blood-myocardium contrast.
2. Increase repetition time (TR) modestly, allowing more time between RF pulses without materially affecting temporal resolution.
3. Use parallel imaging acceleration to reduce the number of RF excitations per unit time.
4. Employ dual-source or multi-transmit RF technology where available, improving RF homogeneity at lower net power.
5. Segment acquisitions per heartbeat conservatively, avoiding unnecessarily high segmentation factors that raise RF duty cycle.

Top 10 pathologies

The diagnostic power of the cardiac MRI protocol lies in its ability to distinguish overlapping clinical presentations — such as chest pain, heart failure, or arrhythmia — into mechanistically distinct disease categories using a shared set of cine, mapping, and LGE sequences. The ten pathologies below are ordered roughly by clinical frequency and represent the conditions most commonly referred for cardiac MRI in general radiology and cardiology practice. Each card summarizes the characteristic T1/T2 relaxation behavior and the specific way the protocol must be tailored to maximize diagnostic yield for that condition.

Pitfalls — radiographers

The primary scanning pitfall for this protocol, as identified in the departmental protocol plan, is ECG misgating caused by arrhythmia, which corrupts the assumption of a regular R-R interval that segmented cine and LGE acquisition depend upon. In segmented acquisitions, data from several heartbeats is interleaved to reconstruct a single representative cardiac cycle; when R-R intervals vary — as in atrial fibrillation, frequent ectopy, or bigeminy — the scanner either misassigns k-space lines to the wrong cardiac phase or triggers prematurely on an ectopic beat, producing visible blurring, ghosting, or a characteristic “sawtooth” discontinuity at the edge of the myocardium.

Pitfalls — radiologists

The primary interpretation pitfall for this protocol is again rooted in ECG misgating from arrhythmia, which can produce blurred wall motion or spurious signal loss that mimics a genuine functional or structural abnormality. Because the resulting blur is often subtle and confined to a single segment, it can be difficult to distinguish from a true regional wall-motion abnormality without deliberately reviewing the corresponding raw ECG trace and cine loop quality — a step that is easy to omit under reporting-volume pressure but essential for diagnostic accuracy.

Pitfalls — non-radiology physicians

Cardiologists, internists, and emergency physicians referring for or acting on cardiac MRI reports rarely need to understand the underlying physics, but misinterpreting a report’s headline findings without appreciating their technical caveats can materially change patient management. The following pitfalls represent the most common gaps between what a report states and what a referring clinician assumes it means.

Pitfall comparison summary

AI and automation

Artificial intelligence has moved from research curiosity to clinical utility in the cardiac MRI protocol, most notably in automated endocardial and epicardial contouring for biventricular volumetry, which reduces inter-observer variability in ejection-fraction calculation. Several deep-learning-based segmentation tools have received FDA clearance and CE marking for automated LV/RV volumetric analysis,^[19][20] and vendor-neutral post-processing platforms now offer automated T1/T2 mapping quality control that flags suboptimal TI nulling or motion-corrupted maps before the reporting radiologist reviews the study.^[18]

Automated arrhythmia-aware reconstruction algorithms are also emerging, using deep-learning-based motion correction to salvage diagnostic-quality cine loops from irregular R-R data that would previously have required a full repeat acquisition — directly addressing the ECG-misgating pitfall that dominates this protocol.^[15] As with all AI tools in radiology, deployment should be limited to evidence-validated, regulatory-cleared software, with human oversight retained for final diagnostic sign-off.

Feature-tracking software applied to standard cine images now allows extraction of global longitudinal, circumferential, and radial strain without any additional acquisition time, adding a sensitive early marker of subclinical dysfunction in conditions such as chemotherapy-related cardiotoxicity and early-stage cardiomyopathy, before ejection fraction itself declines. Departments adopting these tools should validate local reference ranges against their own scanner and reconstruction pipeline, since strain values are known to vary meaningfully between vendors and software versions.

Further reading

1. 7 Proven Strategies for Optimizing MRI Sequences in 2026 — bolus timing and delivery consistency principles directly applicable to cardiac perfusion and LGE imaging.
2. 7 Critical Pituitary Gland MRI Protocol Steps — dynamic multi-phase contrast timing principles relevant to cardiac perfusion sequencing.
3. Acute Stroke MRI Protocol: 10 Critical Steps — perfusion-weighted contrast dynamics and rapid-workflow principles shared with cardiac perfusion CMR.
4. 7 Critical Brain Tumor MRI Protocol Steps — gadolinium dosing and SAR-management strategies transferable to cardiac 3.0T protocols.
5. CTA Aortic Stent Graft Protocol: 5 Critical Phases — cardiovascular contrast-phase timing principles relevant to combined cardiac-aortic imaging workups.

Reducing artefacts with patients and parameters

The most critical scanning parameters that impact image quality in cardiac MRI include the following four domains.

1. Spatial resolution

Spatial resolution defines the ability to distinguish small details in an image. Matrix size (frequency × phase) increases spatial resolution but decreases SNR because voxel size shrinks. Field of view (FOV) reduction increases spatial resolution but similarly reduces SNR through smaller voxels. Slice thickness reduction improves resolution and reduces partial-volume averaging at the endocardial border but significantly decreases SNR — a critical trade-off in thin-walled structures like the RV free wall.

2. Signal-to-noise ratio (SNR)

SNR represents the diagnostic signal relative to background noise. Number of averages (NEX/NSA) improves SNR but roughly doubles scan time per doubling — costly in breath-hold cardiac imaging. Receiver bandwidth reduction boosts SNR but increases scan time and chemical-shift artifact at the fat-myocardium border. Coil selection — dedicated cardiac phased-array coils rather than whole-body coils — captures substantially stronger signal.

3. Image contrast

Repetition time (TR) is the time between RF pulses; short TR maximizes T1 contrast. Echo time (TE) is minimized in bSSFP cine to preserve blood-myocardium contrast and reduce flow-related dephasing. Flip angle controls proton excitation and is a primary determinant of blood-myocardium contrast in gradient-echo-based cine sequences.

4. Artifact control

Phase-encoding direction swaps can shift cardiac and respiratory motion artifact away from the region of interest. Flow compensation and ECG gating use the cardiac trigger to minimize blurring and ghosting from pulsatile motion — the single most important artifact-control mechanism in this entire protocol. Parallel imaging uses multiple coil elements to reduce phase-encoding steps, cutting both scan time and motion sensitivity, which is particularly valuable when breath-hold duration must be minimized in dyspneic cardiac patients.

Parallel imaging protocols and parameters

Parallel imaging acceleration is essential in cardiac MRI to compress cine and LGE acquisitions into breath-holds tolerable by patients with reduced cardiac reserve. The turbo/acceleration factor selected directly trades scan-time reduction against SNR loss and geometry-factor (g-factor) noise amplification.

Conclusion

The cardiac MRI protocol for function and morphology delivers reproducible biventricular volumetry, tissue characterization, and scar detection in a single radiation-free examination, built on ECG-gated segmented cine bSSFP, targeted local shimming, and PSIR late gadolinium enhancement following a 20–30 mL gadolinium bolus. Across the ten pathologies reviewed — from ischemic infarction and cardiomyopathies to amyloidosis, sarcoidosis, and constrictive pericarditis — accurate diagnosis consistently depends on integrating cine function, mapping-based tissue characterization, and LGE pattern recognition rather than any single sequence in isolation.^[5][6][26]

The unifying technical threat across scanning, interpretation, and clinical use of this protocol is ECG misgating in arrhythmia, which can degrade image quality, mimic functional abnormality, and undermine confidence in quantitative results if not actively managed through retrospective gating, real-time cine, or careful lead placement. A disciplined, three-tier pitfall framework — scanning, interpretation, and clinical application — equips radiographers, radiologists, and referring physicians to extract maximum diagnostic value from every cardiac MRI examination.^{[10][12][15][27]}

References

1. Kramer, C. M., Barkhausen, J., Bucciarelli-Ducci, C., Flamm, S. D., Kim, R. J., & Nagel, E. (2020). Standardized cardiovascular magnetic resonance imaging (CMR) protocols: 2020 update. Journal of Cardiovascular Magnetic Resonance, 22(1), 17. https://doi.org/10.1186/s12968-020-00607-1
2. Schulz-Menger, J., Bluemke, D. A., Bremerich, J., Flamm, S. D., Fogel, M. A., Friedrich, M. G., … Kramer, C. M. (2020). Standardized image interpretation and post-processing in cardiovascular magnetic resonance: Society for Cardiovascular Magnetic Resonance (SCMR) recommendations. Journal of Cardiovascular Magnetic Resonance, 22(1), 19. https://doi.org/10.1186/s12968-020-00610-6
3. Messroghli, D. R., Moon, J. C., Ferreira, V. M., Grosse-Wortmann, L., He, T., Kellman, P., … Schulz-Menger, J. (2017). Clinical recommendations for cardiovascular magnetic resonance mapping of T1, T2, T2* and extracellular volume: A consensus statement by the Society for Cardiovascular Magnetic Resonance (SCMR). Journal of Cardiovascular Magnetic Resonance, 19(1), 75. https://doi.org/10.1186/s12968-017-0389-8
4. Ferreira, V. M., Schulz-Menger, J., Holmvang, G., Kramer, C. M., Carbone, I., Sechtem, U., … Friedrich, M. G. (2018). Cardiovascular magnetic resonance in nonischemic myocardial inflammation: Expert recommendations. Journal of the American College of Cardiology, 72(24), 3158–3176. https://doi.org/10.1016/j.jacc.2018.09.072
5. Captur, G., Manisty, C., & Moon, J. C. (2016). Cardiac MRI evaluation of myocardial disease. Heart, 102(18), 1429–1435. https://doi.org/10.1136/heartjnl-2015-308337
6. Salerno, M., Sharif, B., Arheden, H., Kumar, A., Axel, L., Li, D., & Neubauer, S. (2017). Recent advances in cardiovascular magnetic resonance: Techniques and applications. Circulation: Cardiovascular Imaging, 10(6), e003951. https://doi.org/10.1161/CIRCIMAGING.116.003951
7. Puntmann, V. O., Peker, E., Chandrashekhar, Y., & Nagel, E. (2016). T1 mapping in characterizing myocardial disease: A comprehensive review. JACC: Cardiovascular Imaging, 9(1), 67–81. https://doi.org/10.1016/j.jcmg.2015.11.005
8. Arbelo, E., Protonotarios, A., Gimeno, J. R., Arbustini, E., Barriales-Villa, R., Basso, C., … ESC Scientific Document Group. (2023). 2023 ESC Guidelines for the management of cardiomyopathies. European Heart Journal, 44(37), 3503–3626. https://doi.org/10.1093/eurheartj/ehad194
9. Ommen, S. R., Mital, S., Burke, M. A., Day, S. M., Deswal, A., Elliott, P., … Woo, A. (2020). 2020 AHA/ACC guideline for the diagnosis and treatment of patients with hypertrophic cardiomyopathy. Circulation, 142(25), e558–e631. https://doi.org/10.1161/CIR.0000000000000937
10. Kellman, P., & Hansen, M. S. (2017). Myocardial perfusion cardiovascular magnetic resonance: Optimized dual sequence and reconstruction for quantification. Journal of Cardiovascular Magnetic Resonance, 19(1), 43. https://doi.org/10.1186/s12968-017-0355-5
11. American College of Radiology. (2023). ACR Manual on Contrast Media (Version 2023). American College of Radiology. https://www.acr.org/Clinical-Resources/Contrast-Manual
12. American College of Radiology, American Heart Association, & Society for Cardiovascular Magnetic Resonance. (2023). ACR–SCMR practice parameter for the performance and interpretation of cardiac magnetic resonance imaging (MRI). American College of Radiology. https://www.acr.org/Clinical-Resources/Practice-Parameters-and-Technical-Standards
13. International Electrotechnical Commission. (2022). IEC 60601-2-33: Medical electrical equipment — Particular requirements for the basic safety and essential performance of magnetic resonance equipment for medical diagnosis. IEC. https://webstore.iec.ch/publication/64754
14. International Commission on Non-Ionizing Radiation Protection. (2021). Guidelines for limiting exposure to electromagnetic fields (100 kHz to 300 GHz). Health Physics, 118(5), 483–524. https://doi.org/10.1097/HP.0000000000001210
15. Zhuang, B., Sirajuddin, A., Zhao, S., & Lu, M. (2021). The role of 4D flow cardiovascular magnetic resonance in the assessment of normal and diseased cardiac hemodynamics. Quantitative Imaging in Medicine and Surgery, 11(3), 1201–1220. https://doi.org/10.21037/qims-20-873
16. Kellman, P., Wilson, J. R., Xue, H., Ugander, M., & Arai, A. E. (2017). Extracellular volume fraction mapping in the myocardium, part 1: Evaluation of an automated method. Journal of Cardiovascular Magnetic Resonance, 19(1), 43. https://doi.org/10.1186/s12968-016-0295-5
17. Haaf, P., Garg, P., Messroghli, D. R., Broadbent, D. A., Greenwood, J. P., & Plein, S. (2016). Cardiac T1 mapping and extracellular volume (ECV) in clinical practice: A comprehensive review. Journal of Cardiovascular Magnetic Resonance, 18(1), 89. https://doi.org/10.1186/s12968-016-0308-4
18. Bhuva, A. N., Bai, W., Lau, C., Davies, R. H., Ye, Y., Bulluck, H., … Manisty, C. H. (2019). A multicenter, scan-rescan, human and machine learning CMR study to test generalizability and precision in imaging biomarker analysis. Circulation: Cardiovascular Imaging, 12(10), e009214. https://doi.org/10.1161/CIRCIMAGING.119.009214
19. Bai, W., Sinclair, M., Tarroni, G., Oktay, O., Rajchl, M., Vaillant, G., … Rueckert, D. (2018). Automated cardiovascular magnetic resonance image analysis with fully convolutional networks. Journal of Cardiovascular Magnetic Resonance, 20(1), 65. https://doi.org/10.1186/s12968-018-0471-x
20. Chen, C., Qin, C., Qiu, H., Tarroni, G., Duan, J., Bai, W., & Rueckert, D. (2020). Deep learning for cardiac image segmentation: A review. Frontiers in Cardiovascular Medicine, 7, 25. https://doi.org/10.3389/fcvm.2020.00025
21. Petersen, S. E., Aung, N., Sanghvi, M. M., Zemrak, F., Fung, K., Paiva, J. M., … Munroe, P. B. (2017). Reference ranges for cardiac structure and function using cardiovascular magnetic resonance (CMR) in Caucasians from the UK Biobank population cohort. Journal of Cardiovascular Magnetic Resonance, 19(1), 18. https://doi.org/10.1186/s12968-017-0327-9
22. Kellman, P., & Arai, A. E. (2016). Imaging sequences for first pass perfusion — a review. Journal of Cardiovascular Magnetic Resonance, 18(1), 32. https://doi.org/10.1186/s12968-016-0242-5
23. Ferreira, V. M., Piechnik, S. K., Dall’Armellina, E., Karamitsos, T. D., Francis, J. M., Ntusi, N., … Neubauer, S. (2015). T1 mapping for the diagnosis of acute myocarditis using CMR: Comparison to T2-weighted and late gadolinium enhanced imaging. JACC: Cardiovascular Imaging, 6(10), 1048–1058. https://doi.org/10.1016/j.jcmg.2013.03.008
24. Bogaert, J., & Francone, M. (2018). Cardiovascular magnetic resonance in pericardial diseases. Journal of Cardiovascular Magnetic Resonance, 20(1), 14. https://doi.org/10.1186/s12968-018-0439-x
25. Dweck, M. R., Abgral, R., Trivieri, M. G., Robson, P. M., Karakatsanis, N., Mani, V., … Fayad, Z. A. (2018). Hybrid magnetic resonance imaging and positron emission tomography with fluorodeoxyglucose to diagnose active cardiac sarcoidosis. JACC: Cardiovascular Imaging, 11(1), 94–107. https://doi.org/10.1016/j.jcmg.2017.02.021
26. Kramer, C. M. (2019). Role of cardiac MR imaging in cardiomyopathies. Journal of Nuclear Cardiology, 26(3), 833–841. https://doi.org/10.1007/s12350-018-1471-3
27. Fischer, K., Obrist, S. J., Erne, S. A., Stark, A. W., Marggraf, M., Kaneko, K., … Manka, R. (2019). Feature tracking myocardial strain incrementally improves prognostication in myocarditis beyond traditional CMR imaging features. JACC: Cardiovascular Imaging, 13(9), 1891–1901. https://doi.org/10.1016/j.jcmg.2020.04.025