Anak 3

Fratz et al.
Journal of Cardiovascular Magnetic Resonance 2013, 15:51 http://jcmr-

online.com/content/15/1/51
POSITION STATEMENT Open Access
Guidelines and protocols for cardiovascular

magnetic resonance in children and adults with
congenital heart disease: SCMR expert consensus
group on congenital heart disease
Sohrab Fratz1*, Taylor Chung2, Gerald F Greil3, Margaret M Samyn4, Andrew M Taylor5,
Emanuela R Valsangiacomo Buechel6, Shi-Joon Yoo7 and Andrew J Powell8*
Abstract
Cardiovascular magnetic resonance (CMR) has taken on an increasingly important role in the diagnostic evaluation
and pre-procedural planning for patients with congenital heart disease. This article provides guidelines for the
performance of CMR in children and adults with congenital heart disease. The first portion addresses preparation
for the examination and safety issues, the second describes the primary techniques used in an examination, and
the third provides disease-specific protocols. Variations in practice are highlighted and expert consensus
recommendations are provided. Indications and appropriate use criteria for CMR examination are not specifically
addressed.
Keywords: Cardiovascular magnetic resonance, Congenital heart disease, Heart defects, Imaging protocols,
Magnetic resonance imaging
Introduction preparation for the examination and safety issues, the

Over the past two decades, there has been a marked second describes the primary techniques or modules of an
increase in the use of cardiovascular magnetic resonance examination, and the third provides disease-specific
(CMR) for the anatomical and functional evaluation of protocols using these modules. The aim is to provide an
congenital heart disease (CHD) [1-6]. CMR is rarely used educational resource for those engaged in CMR in this
as the initial or sole diagnostic imaging modality. Rather patient population and to help standardize the approach to
it complements echocardiography, provides a non- such patients. As much as possible we try to support our
invasive alternative to x-ray angiography, avoids the recommendations by published evidence. However,
ionizing radiation exposure of computed tomography, and where data is lacking, the recommendations represent an
overcomes many of the limitations of these modalities. expert consensus view.
Expert recommendations regarding the indications for
CMR in adults with CHD have recently been published
Examination preparation and safety
[7] and are underway for children with CHD. This
document focuses on the performance of CMR in children When applicable, parents should be provided with a
and adults with CHD. The first portion addresses detailed description of the CMR examination and asked to
discuss it with their child in an age-appropriate manner in
advance to increase the likelihood of a successful
* Correspondence: fratz@dhm.mhn.de;
andrew.powell@cardio.chboston.org
1 Department of Pediatric Cardiology and Congenital Heart
Disease,
Deutsches Herzzentrum München (German Heart Center Munich) of the
Technical University Munich, Munich, Germany
8 ’s Hospital, and the Department
Department of Cardiology, Boston Children of
Pediatrics, Harvard Medical School, Boston, MA, USA
Full list of author information is available at the end of the article
Fratz et al. Journal of Cardiovascular Magnetic Resonance 2013, 15:51 Page 2 of 31http://jcmr-online.com/content/15/1/51
study. Prior to bringing the patient into the scanner room, the physician and technologists should review the patient’s
history and safety screening form to identify implanted devices which may be hazardous in the CMR environment or
produce image artifact. For patients who may have undergone a cardiac procedure and have an
© 2013 Fratz et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution,
and reproduction in any medium, provided the original work is properly cited.
incomplete or unreliable history, a chest radiograph a feeding [11,12]. The baby is undressed, prepared with
should be obtained to assist screening. A detailed ECG-leads and an oxygen saturation probe, and fed. As
discussion on CMR safety and device interactions can be soon as the infant starts falling asleep, it is swaddled or
found elsewhere [8-10]. wrapped in an immobilizer, provided noise protection,
Following safety screening, physiological monitoring and placed on the scanner table. With this “feed, swaddle,
devices and hearing protection (for both awake and and sleep” technique, scanning times of 30–60 minutes
sedated patients) are put in place. Young children can be achieved. However, this approach does not allow
dissipate body heat faster than adults; thus, patient suspension of respiration to reduce motion artifact. As the
temperature should be monitored and blankets applied as study can be compromised by early awakening, the
needed to minimize heat loss. A high-quality imaging protocol should be tailored strictly to the highest
electrocardiogram (ECG) signal is essential for optimum priority clinical questions.
data quality in cardiac-gated sequences. The adequacy of Alternatively, deep sedation can be achieved with a
the signal should be checked not only at the onset when variety of medications (e.g., pentobarbital, propofol,
the patient is outside the scanner bore, but also once fentanyl, midazolam, and inhalational agents). Meticulous
inside it and during actual scanning. In patients with care must be taken to maintain spontaneous breathing
dextrocardia, ECG leads are best placed on the right chest. under the supervision of an experienced anesthesiology
The imaging coil should be chosen to maximize the team. The principal drawbacks of this approach are an
signal-to-noise ratio over the body region to be examined. unprotected airway and reliance on spontaneous
Because CHD often involves abnormalities of the thoracic respiratory effort with the associated risks of aspiration,
vasculature, the coil will usually need to be large enough airway obstruction, and hypoventilation. A laryngeal
to cover the entire thorax and upper abdomen rather than mask airway may be used in conjunction with deep
just the heart. In smaller patients, pediatric thoracic coils sedation to decrease the risk of aspiration. From an image
or coils designed for an adultsized head, shoulder, or knee quality standpoint, respiratory motion artifact may reduce
may be suitable. Adequate coil coverage and placement sharpness. However, images acquired from sedated,
should be confirmed early in the examination by freely-breathing patients are often still of diagnostic
reviewing the localizing images. quality as the breathing pattern tends to be particularly
consistent and thus quite amenable to commonly available
Sedation respiratory motion compensation techniques.
Patients undergoing CMR must remain still in the scanner Because of these safety and image quality concerns, some
for up to 60 minutes to minimize motion artifact during institutions prefer to perform sedation using general
image acquisition and allow planning of successive anesthesia with mechanical ventilation and endotracheal
imaging sequences. Young children (typically less than intubation. This approach consistently achieves adequate
age 6 to 8 years) and cognitively impaired older patients sedation, protects the airway, and offers control of
typically require some form of sedation. Multiple factors ventilation. Respiratory motion artifact can be eliminated
should be taken into account when deciding whether a by suspending ventilation for brief periods (15–60
patient should have an examination with sedation seconds) in conjunction with neuromuscular blockade.
including the length of the anticipated examination Compared to deep sedation, general anesthesia tends to
protocol, developmental maturity, the patient’s experience require more specialized personnel and greater equipment
with prior procedures, the parents’ opinion of their child’s resources (e.g., a magnetic resonance (MR) compatible
capability to cooperate with the examination, the risks of anesthesia machine). Both the deep sedation and general
sedation, and the benefits of the diagnostic information. anesthesia strategies for CMR have been shown to have a
Strategies for sedation and anesthesia during CMR vary good safety profile in this fragile patient population [13-
and often depend on institutional preference and 18]. MR compatible equipment should be used to monitor
availability of resources such as pediatric the heart rate, transcutaneous oxygen saturation, blood
anesthesiologists with experience in CHD. Sedation in pressure, expired carbon dioxide, and body temperature.
infants younger than approximately six months can be An appropriately equipped resuscitation cart and
achieved by allowing them to fall into a natural sleep after emergency management plan for the MR environment
should be in place. To maximize patient safety and relaxation time (T1) at 3 T, typically lead to improved
examination quality, it is recommended that different clinical performance of coronary angiography,
healthcare providers be responsible for supervising the contrastenhanced angiography, myocardial tagging, and
imaging and sedation/anesthesia aspects of the study, and myocardial perfusion imaging sequences. However, MR
that both communicate closely with each other. imaging at 3 T has inherently stronger off-resonance
artifacts (B0-field inhomogeneity) and dielectric shading
Gadolinium-based contrast agents artifacts (B1-field inhomogeneity). These factors translate
Gadolinium-based intravenous MR contrast agents to significant dark band artifacts and loss of tissue
(GBCA) are commonly administered in CHD patients of contrast on steady-state free precession (SSFP) pulse
all ages for angiography and the assessment of myocardial sequences and signal loss on spin echo pulse sequences
perfusion and viability. These agents are often used “off- that are unpredictable [26]. Strategies to reduce some of
label” in children as several of them are not approved by these artifacts are currently under evaluation [27]. It is
regulatory agencies, such as the United States Food and also worth noting that implanted metallic devices (e.g.,
Drug Administration or the European Medicines Agency, sternal wires, stents, septal occluders, and vascular
for pediatric-age patients. The incidence of adverse events occlusion coils) are commonly encountered in CHD
related to GBCA in both adults and children is very low patients referred for CMR [28]. Device-related signal loss
[14,19-21]. The vast majority of these reactions are mild artifact from T2* effects is typically more pronounced at
and include coldness, warmth, or pain at the injection site; higher field strengths. Moreover, CMR compatibility
nausea; vomiting; headache; paresthesias; dizziness; and information for devices is more commonly available at the
itching. Severe, life-threatening anaphylactoid or 1.5 T than the 3 T field strength.
nonallergic anaphylactic reactions are quite rare (0.001%
to 0.01%) [22]. There is no evidence indicating Common CMR techniques (modules) in CHD
nephrotoxicity at approved dosages. This section focuses only on established CMR techniques
GBCA administration to patients with acute renal failure which are in routine use throughout the CMR community
or severe chronic kidney disease is associated with the and are available from all scanner manufacturers. The
development of nephrogenic systemic fibrosis (NSF), a authors readily acknowledge that newer applications such
rare and serious condition that involves fibrosis of as 3D phase-contrast flow measurement, real-time and 3D
primarily the skin and subcutaneous tissue but may also cine imaging, and blood pool contrast agents can be useful
involve the lungs, esophagus, heart, and skeletal muscles. when performing CMR in CHD patients.
Patients with an estimated glomerular filtration rate <30
ml/min/1.73 m2 are considered at highest risk. Thus, all Spin echo
patients who are candidates for GBCA administration Spin echo pulse sequences are typically used in CMR to
should be screened for renal dysfunction, and, if generate images in which flowing blood appears dark and
identified, the most recent institutional or national more stationary tissues appear as varying shades of gray
guidelines regarding GBCA use should be consulted or white (Figure 1). The most common variants of this
[22,23]. There are only a small number of reported cases technique are fast (turbo) spin echo (commercial names:
of NSF in children (fewer than 20 as of 2010), the TSE, Siemens; FSE, General Electric; TSE, Philips) and
youngest of which was 8 years of age [24,25], and all had single-shot fast (turbo) spin echo (commercial names:
significant renal dysfunction [22]. No NSF cases have HASTE, Siemens; SSFSE, General Electric; SSTSE,
been reported in preterm or term neonates despite their Philips). Both versions usually employ ECGtriggering to
immature kidney function and estimated glomerular compensate for cardiac motion and preparation pulses to
filtration rates which may be < 30 ml/min/1.73 m 2 [25]. suppress signal from blood and improve image contrast.
Accordingly, caution and a careful assessment of benefits Respiratory motion can be addressed by breath-holding,
and risk but not prohibition are advised when acquiring multiple signal averages, triggering from
administering GBCA to neonates and infants [22,23]. respiratory bellows, or respiratory-navigator gating. By
utilizing half-Fourier k-space filling, the single-shot
Scanner field strength technique acquires all the data to make an image in one
Most CMR is performed on 1.5 T or 3 T scanners. In heartbeat and is thus very time efficient. The fast spin
general, 3 T yields a higher signal-to-noise ratio, and echo technique collects data for one image
therefore allows for better spatial resolution, which is
particularly desirable in younger, smaller patients. This
stronger signal and the inherently longer longitudinal
For rates greater than 100 bpm, it may be helpful to

compensate for this effect by acquiring image data every
third or fourth heartbeat. Higher heart rates are also
accompanied by faster cardiac motion. With the fast spin
echo sequence, the number of echoes during the readout
(echo train length) can be decreased to reduce the image
acquisition (shot) duration and better resolve rapidly
moving structures. With some implementations of spin
echo imaging, the blood suppression preparation pulse
may become ineffective at faster heart rates and blood
will appear brighter. Finally, the default blood
Figure 1 Spin echo and gradient echo cine.63-year-old patient suppression preparation pulse will not be effective
who has undergone surgical repair of coarctation of the aorta. The
images are oriented parallel to the long-axis of the aortic arch and
following the administration of GBCA; thus, in most
shown in diastole. A. ECG-triggered fast spin echo sequence protocols evaluating anatomy, spin echo sequences should
acquired with a double inversion preparation pulse to suppress be performed prior to contrast agent administration.
signal from flowing blood. Note that the resulting signal from blood
is dark. B. ECG-gated steady-state free precession cine sequence. Gradient echo cine
Note the signal from blood is bright.
Gradient echo cine pulse sequences generate images in
over multiple heartbeats and thus takes longer than the which flowing blood appears bright (Figure 1). With the
single-shot technique, but it produces higher resolution use of ECG-gating, it produces multiple images across the
images. cardiac cycle that can be displayed in cine loop format to
In contrast to cine gradient echo techniques, spin echo visualize motion, one of its main advantages over spin
images are usually acquired during only one portion of the echo sequences. Respiratory motion artifact can be
cardiac cycle and do not depict motion; they are thus minimized by breath-holding (preferred when possible) or
typically used to provide anatomic information. Their by acquiring 2–4 signal averages with the patient
main advantage over cine imaging is that they are less breathing. In clinical practice, this imaging sequence is
susceptible to artifacts caused by turbulent flow and often prescribed across the anatomy of interest to yield a
metallic implants such as sternal wires, septal occluders, stack of contiguous cross-sectional slices that can be
stents, and vascular occlusion coils (Figure 2) [28]. In displayed in a multi-location, multi-phase format.
addition, a thinner slice thickness (≈2 mm) can be Table 1 Fast (turbo) spin echo
achieved which may be particularly useful in smaller Infant/small child
patients. Fast spin echo sequences can also be modified to
alter image contrast (e.g., T1- and T2-weighting, and fat In-plane resolution (mm) 1.0-2.0
suppression) and thereby help characterize tissue
composition. Slice thickness (mm) 2-3
For smaller patients, the operator should use a smaller
voxel size (adjusting both in-plane resolution and slice Echo train length 12-24
thickness) to ensure adequate spatial resolution (Table 1).
If needed, the reduction in signal-to-noise ratio that results Image acquisition timing 3-4 R-R
may be offset by employing multiple signal averages.
Respiratory compensation Free-breathing
Data acquisition should be timed to the portion of the
cardiac cycle in which the heart has the least motion (i.e.,
Number of signal averages 3
the rest period) to minimize blurring. At lower heart rates
the rest period is often in mid-diastole; whereas at higher Trigger delay Diastole or systole
heart rates (> 90-100 bpm) it may be at end-systole. Fast
spin echo sequences designed for heart rates seen in adults
usually acquire data every heartbeat or every second
heartbeat and thus the repetition time (TR) is equal to 1 R-
R or 2 R-R intervals, respectively. At higher heart rates,
the R-R interval becomes shorter and the time between
data acquisition will decrease resulting in less time for
longitudinal signal recovery and thus lower image quality.
Gradient echo cine imaging can be performed using a vasculature in a short scan time, typically less than 30
standard spoiled gradient echo pulse sequence or the seconds (Figure 3). As CHD
subsequently developed SSFP sequence (commercial
names: TrueFISP, Siemens; FIESTA, General Electric;
Infant/small c
balanced-FFE, Philips). SSFP imaging is faster and
provides superior contrast between blood and In-plane resolution (mm) 1.2-2.0
myocardium compared to standard gradient echo imaging
and is thus more commonly used. The SSFP sequence is Slice thickness (mm) 4-6
also relatively less sensitive to flow disturbances caused
Inter slice gap (mm) 0-2
Respiratory compensation Free-breathing
Number of signal averages 3
Reconstructed phases per R-R interval
ECG gating
Figure 2 Stainless steel coil artifact.17-year-old patient with dextrocardia, double-outlet right ventricle, and pulmonary stenosis who has
undergone a Rastelli operation and catheter implantation of a single stainless steel vascular coil to occlude a small left superior vena cava
draining to the left atrium. ECG-gated steady-state free precession sequence in a coronal planeA)( and axial plane (B) demonstrating a region of
signal loss in the upper left chest (arrow) resulting from the coil. The area of signal loss is several times larger than the coil and obscures a
portion of the left pulmonary artery. C( ) shows an ECG-triggered fast spin echo sequence with a double inversion preparation pulse acquired in
diastole with a similar orientation as in B( ) With this sequence, the extent of signal loss artifact is reduced, and the left pulmonary artery is more
clearly visualized.
by stenotic or regurgitant jets. However, it is also more Table 2 Cine steady-state free precession
prone to image artifact when there are inhomogeneities in is frequently associated with abnormalities of the chest
the magnetic field (B0) caused by suboptimal scanner vasculature, this technique is often employed in both pre-
shimming or implanted ferromagnetic devices. Thus, the and post-operative CMR imaging protocols. Studies
standard gradient echo technique may be preferred for demonstrating its utility and accuracy in patients with
visualization of flow jets or imaging near implanted CHD have been published for examination of the aorta
devices such as stents and mechanical valves. and its branches, pulmonary arteries, pulmonary veins,
When imaging smaller structures in younger patients, the systemic veins, aortopulmonary and venousvenous
spatial resolution should be increased appropriately (Table collateral vessels, systemic-to-pulmonary artery shunts,
2). Changes to accomplish this (e.g., increased matrix conduits, and vascular grafts [29-34]. The 3D datasets
size, smaller field of view, thinner slice thickness), generated by MRA are well-suited to volume rendered
however, will prolong the echo time (TE). Once the TE displays which can enhance understanding of complex
becomes > 2 ms and the TR > 4 ms, the quality of SSFP spatial relationships and are more comprehensible to non-
imaging often deteriorates. One can increase the matrix CMR specialists (Figure 3). Nevertheless, it is essential
size only in the phase encode direction to improve that the source data be carefully reviewed by the reporting
resolution without prolonging the TE, but this is done at physician as anatomic information may be omitted or
the expense of increased acquisition time. Thus, a careful distorted by the volume rendering algorithm.
balance must be struck between spatial resolution, When possible, MRA should be performed with breath-
acquisition time, and image quality. Alternatively, some holding to minimize artifact from respiratory motion [35].
centers prefer to use high-resolution standard gradient In smaller patients, one should ensure that spatial
echo imaging with free-breathing and multiple signal resolution, both in-plane and partition thickness, is
averages. sufficient. A contrast agent dose of 0.1-0.2 mmol/kg is
typically used. The time delay between the start of
Contrast-enhanced magnetic resonance angiography Magnetic contrast injection and data acquisition will influence
resonance angiography (MRA) using an intravenously which portions of the vasculature will be depicted. This
administered GBCA can produce a high-resolution, high- interval can be determined by a MR fluoroscopic method
contrast three-dimensional (3D) dataset of the entire chest that allows real-time visualization of the arrival of the
contrast bolus to the targeted region, by a preceding Another disadvantage is that the use of GBCA incurs
timing run with a smaller dose of contrast, or by some risk of adverse reactions (see above).
automatic bolus detection. Because the effect of the ECG and respiratory navigator-gated 3D SSFP
contrast agent In its typical implementation, the ECG and respiratory
navigator-gated 3D SSFP technique delivers a 3D
anatomic dataset with an isotropic voxel size of
approximately 1.2-2.0 mm without the use of a contrast
agent (Table 3). Its utility and validation have been
reported in patients with CHD [40-42]. Using ECG
triggering, the data acquisition is confined to one or two
portions of the cardiac cycle thereby minimizing blurring
from cardiac motion. Intracardiac anatomy and coronary
arteries can thus be more clearly visualized than with
contrastenhanced MRA. The 3D SSFP sequence is
performed with free-breathing. Respiratory motion is
compensated for by gating data acquisition to expiration
Figure 3 Contrast-enhanced magnetic resonance angiography.
with the use of a navigator beam tracking the motion of
9-year-old patient with partially anomalous pulmonary venous return
of the left upper pulmonary vein (arrow) to the leftward aspect of the diaphragm. This approach allows for improvement in
the left innominate vein. Contrast-enhanced magnetic resonance spatial resolution including isotropic voxel size because
angiogram shown in a coronal plane using a sub-volume maximal scan time is not limited to the duration of a single breath-
intensity projection (A) and volume rendering (B). hold. The isotropic property of the anatomic data permits
arbitrary reformatting in any desired imaging plane during
review without the loss of resolution (Figure 4).
persists even during recirculation, two or three sequential
The 3D SSFP sequence has four principal disadvantages.
MRA data acquisitions are useful to ensure visualization
First, its acquisition time is relatively long, usually
of all the vasculature. More recently, imaging acceleration
approximately 7–10 min, during which the patient must
techniques have been applied to shorten the acquisition
be absolutely still. Younger children may have difficulty
time to 2–5 seconds thereby permitting multiple 3D
with this level of cooperation. Second, the sequence is
volume sets to be acquired as the contrast agent passes
very susceptible to artifacts caused by turbulent flow and
through the circulation producing a “time-resolved” MRA
magnetic field inhomogeneity such as that caused by
[36-39]. A smaller contrast dose (0.05-0.1 mmol/kg) is presence of stents or other ferromagnetic implants.
often used. This time-resolved approach has several Therefore, vessels with stenosis or regurgitation, and
potential advantages: 1) observing the passage of contrast structures around stents can sometimes be misinterpreted.
may have diagnostic benefits, 2) timing of the acquisition Third, the data is confined to one or two portions of the
is less critical because multiple volume sets are obtained cardiac cycle thereby precluding the evaluation of cardiac
as the contrast passes through the circulation, and 3) it and vessel motion. Finally, as with other ECG-gated
may have less sensitivity to respiratory motion artifact. techniques, image quality will suffer when the patient has
The principal disadvantage with this technique is that the an irregular heart rhythm. In such cases, generating stacks
decrease in acquisition time is usually achieved in part of thin SSFP localizer type images in the cardinal planes
from a reduction in spatial resolution which may hinder
during a breath-hold or free-breathing with 2–3 signal
diagnostic accuracy, particularly in smaller patients. In
averages may be a useful alternative.
addition, the undersampling of k-space used to accelerate
A version of 3D SSFP is generally the technique of choice
the acquisition may result in image artifact.
for coronary artery imaging in patients with CHD (Figure
Both the standard and time-resolved contrast MRA 5) [43-46]. For this purpose in particular, data acquisition
techniques do not utilize ECG-gating; therefore, motion must be confined to the rest period of the cardiac cycle
over the cardiac cycle causes blurring, particularly of the (i.e., that with the least motion) to minimize blurring of
aortic root, coronary arteries, and intracardiac structures. these small, fast moving structures.
Table 3 ECG and respiratory navigator-gated 3D steady-state free precession
Infant/small child Large child/adult
Isotropic resolution (mm3) 1.2-1.5 1.3-2.0

Navigator window (mm) 3 5
Image acquisition duration (ms) 40-60 80-150

Trigger delay End-systole or mid-diastole Mid-diastole
The rest period of the heart is chosen by reviewing a high evaluation of the right ventricle or single ventricle. It is a
temporal resolution cine image of the heart (≥50 images key component of the CMR functional evaluation in
per cardiac cycle), usually a 4-chamber view, and patients with CHD.
identifying the appropriate trigger delay and acquisition
(shot) duration. Younger patients will typically have faster
heart rates and thus require a shorter acquisition duration.
Moreover, at higher heart rates (> 90-100 bpm) the
optimal rest period may be in end-systole. If a patient has
difficulty being still during the scan and the primary
diagnostic goal is to visualize the proximal
Figure 5 Coronary MR angiography. Patient with anomalous

origin of the
Figure 4 3D steady-state freeright coronary artery
precession.A from
patient thetransposition
with left aortic sinus of great arteries who has undergone a Senning operation. An ECG
of the
Valsalva. Coronary
and respiratory navigator-gated angiography
3D SSFP sequencewaswasperformed
utilized tousing an ECG
generate andvolume with 1.5 mm isotropic resolution timed to mid-diastole.
a 3D
respiratory navigator-gated 3D SSFP sequence with data acquisition
The navigator efficiency was 45% and the acquisition time was 6 minutes. Multiplanar reformatting of this volume allows a comprehensive
timed to
morphologic evaluation of the
the diastolic restgreat
heart and period of the
vessels cardiac the
including cycle. Multiplanar
Senning pathwaysA,( B, and C).
reformatting oriented in short-axis to the aortic root yielded
this image.
coronary arteries (e.g., suspect anomalous origin of a
coronary artery), it may be helpful to shorten the scan
duration by prescribing a smaller, more targeted imaging
volume around the aortic root rather than the entire heart.
Ventriculography
CMR ventriculography generates cine images of the
ventricles that enable calculation of ventricular volume,
mass, and ejection fraction as well as evaluation of
regional wall motion. It is a specialized application of
gradient echo cine imaging described above. CMR is
widely considered the clinical reference standard for
ventriculography. It is particularly useful in situations
where echocardiography has significant limitations such
as in patients with poor acoustic windows and the
In addition to the technical points mentioned in the

gradient echo cine imaging section, the use of
retrospective rather than prospective ECG-gating is
important so that the entire diastolic portion of the cardiac
cycle is evaluated. When possible, breath-hold
acquisitions are recommended, preferably at end-
expiration as diaphragmatic position tends to be more
consistent from breath-hold to breath-hold [47,48].
Parallel imaging or partial Fourier techniques can be
utilized to decrease breath-hold time but at the expense of
image quality (e.g., poorer signal-to-noise ratio).
Alternatively, freebreathing imaging with the patient
instructed to breathe regularly and quietly, and multiple
signal averaging (2–4) can be performed. Proper attention
to prescribing an appropriate temporal resolution is
essential to adequately depict cardiac motion and capture
the end-systolic phase of the cardiac cycle. Cine CMR
sequences often employ view sharing (also called echo
sharing) to increase the apparent temporal resolution by
undersampling k-space data for certain frames and
utilizing data from adjacent frames to fill the missing data
[49]. Although these interpolated frames smooth cardiac
motion, they do not increase the true temporal resolution
of the sequence which can be calculated as the product of
the TR and lines per segment (also called views per
segment or turbo field echo (TFE) factor). It is
recommended that a minimum of 15 noninterpolated
images be acquired over the cardiac cycle (R-R Figure 6 Ventricular views. The diagram illustrates one approach
interval/(TR•lines per segment) ≥ 15). At the higher heart to planning standard ventricular views (right column) based on
adjusting the slice locationon two other views (left and middle
rates typically seen in younger children, this will require columns). Note that for the 4C, LV 2C, and LV 3C, the imaging plane
that the number of lines per segment be decreased to is carefully positioned to pass through the apex of the LV and bisect
maintain adequate temporal resolution. the mitral valve plane. LV, left ventricle; RV, right ventricle; RVOT,
Imaging planes for ventriculography should be planned right ventricular outflow tract; SA, short-axis; 2C, 2-chamber; 3C, 3-
chamber; 4C, 4-chamber.
carefully and include the following (Figure 6): 1) 4-
chamber (horizontal long-axis) view, 2) left ventricular 2- outflow tracts, 4) right ventricular 3-chamber (right
chamber (vertical long-axis) view, 3) left ventricular 3- anterior oblique) view including tricuspid valve inflow
chamber view including the mitral valve inflow and left and right ventricular outflow, and 5) a stack of contiguous
ventricular slices which completely encompass both ventricles
(Figure 7). The data from the ventricular stack are used to
calculate ventricular end-diastolic volume, end-systolic
volume, stroke volume, ejection fraction, and myocardial
mass using the summation of disks method. If the patient
is expected to have limited tolerance for the examination,
acquiring the ventricular stack should be prioritized.
There is practice variation regarding the orientation of the
ventricular stack. Some centers prefer to prescribe it in a
ventricular short-axis plane parallel to the atrioventricular
valve plane or perpendicular to the ventricular septum
(assisted by cross-referencing with the long-axis images),
and others prefer to prescribe an axial plane (Figure 7)
[50-55]. Furthermore, some advocate acquiring data sets
in both the short-axis and axial planes, or two sets of

short-axis planes—one oriented parallel to the mitral valve
plane and one oriented parallel to the tricuspid valve plane
as these may be offset, particularly when there is right
ventricular dilation. The principal advantages of the axial
plane over the short-axis plane include 1) straightforward
planning especially in patients with complex ventricular
morphology, 2) easier identification of the atrioventricular
boundary during post-processing, and 3) more anatomic
coverage of non-ventricular structures (e.g., atria). Among
its disadvantages are 1) difficulty in evaluating left
ventricular segmental wall motion using established
guidelines, and 2) a limited evaluation of ventricular mass
because the epicardial and endocardial borders of the
diaphragmatic wall of the heart are not clearly defined.
With either approach, it is essential that the entire
ventricle be included in the image stack. For the short-axis
orientation, crossreferencing the slices on the 4-chamber
view may reveal the need to extend the stack above the
tricuspid valve plane in patients with a dilated right Figure 7 Planning for ventriculography. An axial stack of cine images for ventriculog
ventricle (Figure 7). coronal and sagittal images (top row). A short-axis stack of cine images for ventriculog
The ventricular image stack is typically analyzed by chamber (4C) and left ventricular 2-chamber (LV 2C) images in diastole (bottom row). N
demarcating endocardial and epicardial borders with the to ensure complete coverage of the left and right ventricles. In this short-axis example,
septum on the 4C view, and care is taken to ensure that coverage includes the anterio
assistance of software tools (Figure 8). Cross-referencing above the tricuspid valve plane. An alternative short-axis planning approach is to orient
the ventricular short-axis images with long-axis images the 4C view (not shown).
and observing wall motion in a cine fashion facilitate
accurate determination of the atrioventricular and
semilunar valves planes [56,57]. In the absence of right
ventricular hypertrophy, the epicardial boundary of the
thin-walled right ventricle can be difficult to detect. When
the left ventricle is in its normal systemic position, the
mass of the ventricular septum is counted as part of the
total left ventricular mass. When the right ventricle is in
the systemic position, there is no consensus regarding to
which ventricle the septal mass should
Figure 8 Tracing ventricular borders.Diagram demonstrating the drawing of left and right ventricular endocardial contours in end-diastole.
Images may be acquired in a ventricular short-axis orientation A( and B) or in an axial orientation(C and D). There is practice variation regarding
whether to trace the papillary muscles and right ventricular trabeculations to exclude them fromA( and C) versus include them in (B and D) the
blood pool.
be allocated. In patients with ventricular conduction A PC CMR pulse sequence produces two sets of cine
delay, which is relatively common in CHD, the images: magnitude images that provide anatomic
endsystolic and end-diastolic frames (i.e., time point in information and phase images in which the velocity
the cardiac cycle) may not be the same for the right and information is encoded. On the phase images, the signal
left ventricles and should thus be selected independently amplitude (brightness) of each voxel is proportional to
to yield the minimum and maximum volumes mean flow velocity within that voxel. Maximum velocity
respectively. There is practice variation regarding whether in one direction is displayed as the brightest white,
to trace the papillary muscles and right ventricular maximum velocity in the opposite direction as the darkest
trabeculations in order to exclude them from blood pool black, and zero velocity as mid-grey. Using specialized
and count them toward ventricular mass (Figure 8) [58- software, regions of interest around a vessel are defined,
61]. Tracing these structures will yield smaller ventricular and the flow rate is automatically calculated from the
volumes, with little change in stroke volume but a higher product of the mean velocity and the cross-sectional area.
ejection fraction. Though theoretically more accurate, this Blood flow measurements are an important element of the
approach is time consuming in the absence of a reliable CMR examination of CHD patients and have variety of
automated system and may reduce measurement applications. Examples include measurement of cardiac
reproducibility. output [74,75], pulmonary-to-systemic flow ratio (Q p/Qs)
Normative data on ventricular parameters by CMR are [76-81], differential lung perfusion [82-85], valvular
available for both adults and older children regurgitation [55,86-94], aortopulmonary collateral flow
[50,52,53,62,63]; there is still a need for robust data in [95-97], and pressure gradient [98-101].
patients younger than 8 years. Given the practice For PC CMR flow measurements, the use of retrospective
variations in the methods of acquiring and analyzing the rather than prospective ECG-gating is preferred so that the
ventriculography image data noted above, it is entire diastolic portion of the cardiac cycle is evaluated
recommended that a center’s approach to these issues (Table 4). The quality of the ECG-gating should be
correspond to that used in the normative dataset that has carefully monitored, particularly during acquisitions
been selected as the center’s reference values. Although lasting several minutes. If the heart rate changes
indexing ventricular volumes to body surface area is a significantly or there are frequent invalid triggers, the
common practice, this does not fully account for changes sequence should be stopped and repeated. Measurements
in body size from infancy to adulthood as the volumes do can be performed with free-breathing and multiple signal
not vary linearly with body surface area averages (2–4) or by shortening the scan time so that it
[52,64,65]. Thus, a given indexed volume (e.g., end- can comfortably be performed within a breath-hold. The
diastolic volume of 80 ml/m2) may be normal for an adult reduced scan time is usually achieved by reductions in
but above normal for an infant. temporal and spatial resolution, as well as faster gradient
One of the strengths of CMR ventriculography is that switching–all of which may increase measurement error.
studies have demonstrated very good reproducibility in Moreover, physiological changes can occur with breath-
children and adults with CHD [58,66-69]. To achieve this holding that alter the measurements [102-104] and may
level of quality and reliability, centers should maintain a confound clinical interpretation. For these reasons, the
rigorous and consistent approach to imaging and analysis. authors recommend free-breathing acquisitions. The scan
Furthermore, to optimize interstudy reproducibility in time of these measurements may be reduced with the use
patients followed longitudinally, the ventricular borders of parallel imaging [80,105,106].
demarcated in the analysis software should be saved so Table 4 Velocity-encoded phase-contrast cine for quantitative flow measuremen
that they can be compared side-by-side with those from Infant/small child
subsequent studies.
In-plane resolution (mm) 1.0-1.3
Blood velocity and flow measurement Slice thickness (mm) 5
Velocity-encoded phase-contrast (PC) cine CMR is the Number of signal averages
primary technique used to measure blood flow velocity
and volume. A detailed description can be found Reconstructed phases per R-R interval
elsewhere [70-73]. In brief, PC CMR is based on the
principle that the signal from hydrogen nuclei (such as Velocity encoding (cm/s) Artery
those in blood) flowing through specially designed Cardiac/respiratory motion Ret
magnetic field gradients accumulates a predictable and Accurate PC CMR measurements require sufficient
measurable phase shift that is proportional to its velocity. spatial resolution to avoid significant partial volume
effects. Specifically, there should be more than 3 pixels vessel’s or valve’s orientation using two orthogonal
across the diameter or more than 8 pixels in the planning views in order to minimize partial volume
crosssection of the vessel or cardiac valve of interest effects (Figure 10), and the velocity encoding direction
[107,108]. Proper attention to prescribing an appropriate should be set to the through-plane direction. The vessel of
temporal resolution is also essential because under- interest should be as close to the scanner’s isocenter as
sampling will smooth a pulsatile flow curve and cause possible to maximize gradient fidelity (Figure 11). This is
inaccuracies. For PC CMR, true temporal resolution (as accomplished by prescribing the plane so that the center
opposed to interpolated temporal resolution) equals of the image is at the same level as the vessel in the
2•TR•lines per segment (lines per segment is also called superoinferior direction. The scanner will then slide the
views per segment or TFE factor). It is recommended that patient table so that the center of the imaging plane and
a minimum of 20 noninterpolated images be acquired over thus the vessel is close to the scanner’s isocenter.
the cardiac cycle (R-R interval/(2•TR•lines per segment) Placement of the imaging plane in regions of turbulent
≥ 20). At the higher heart rates typically seen in younger flow should be avoided as such flow can lead to signal
children, this will require that the number of lines per loss and inaccuracies. Similarly, positioning of the
segment be decreased to maintain adequate temporal imaging plane away from ferromagnetic implanted
resolution. Finally, with PC CMR measurements, the devices is recommended as these disrupt the magnetic
operator must set the velocity range (v enc) prior to running field gradients.
the pulse sequence; velocities which exceed this range There is some uncertainty and practice variation regarding
will alias and be misrepresented (Figure 9). It is the optimal location for semilunar valve flow
recommended that the venc be set to approximately 25%
above the expected maximum velocity so as to optimize
the dynamic range. When blood velocity is increased,
such as in the setting of valve or vessel stenosis, the
operator’s choice of venc may be guided by information
from recent echocardiography. If aliasing occurs, the
measurement should be repeated using a higher v enc or the
data may be rescaled with cautious use of post-processing
software.
Careful positioning of the PC CMR imaging plane is
essential for accurate velocity and flow measurements. It
should be aligned strictly perpendicular to the blood
Figure 9 Effect of aliasing on phase-contrast cine CMR (PC CMR) flow measurements.Sixteen-year-old patient with surgically repaired
tetralogy of Fallot and mild pulmonary valve stenosis. PC CMR was performed in the main pulmonary artery with the velocity rangeenc (v) set
incorrectly at 200 cm/sec (top row) and then with the evnc set correctly at 300 cm/sec (bottom row). Magnitude (A, D) and phase images (B, E)
in systole, and the resulting flow curves C,( F) generated from analyzing the region of interest (yellow contour) are shown. Because the peak
velocity is 260 cm/sec, aliasing B
( ) and flow underestimation (C) are seen with a venc of 200 cm/sec but not with a venc of 300 cm/sec (E and F).
measurement. A position image. Again, through-

closer to the annulus has plane motion will
the potential advantages of compromise the accuracy
a less complex flow of flow measurements
pattern, reduced impact of made with conventional
vessel compliance, and 2D PC CMR
exclusion of coronary flow [111,118,119].
in the case of the aortic PC CMR has been utilized
valve [109]. Through-plane to estimate the pressure
motion, however, is drop across discrete
greatest near the annulus, stenoses in valves and
and this contributes to vessels by measuring the
measurement error maximum flow velocity
[110,111] and precludes and applying the modified
precise placement between Bernoulli equation. A
the annulus and coronary useful way to plan this is to
ostia throughout the prescribe a PC CMR
cardiac cycle using imaging plane parallel to
standard two-dimensional the direction of blood flow
(2D) PC CMR. Most setting the velocity
operators typically place encoding direction in the
the imaging plane at the same in-plane direction,
Figure 10 Planning a CMR phase-contrast acquisition to measure flow in the main pulmonary
sinotubular junction artery.
orThe andPC CMR imaging
then to plane is
prescribe a
simultaneously viewed and adjusted on orthogonal views of the main pulmonary artery (top row) to ensure that it is oriented perpendicular to
proximal
the blood vessel. The resulting magnitude and phase images are shown (bottom row).
ascending aorta second PC CMR
to measure aortic valve acquisition with through-
flow, and in the middle of plane encoding
the main pulmonary artery perpendicular to the
to measure pulmonary location of the peak
valve flow [112]. inplane velocity. However,
Atrioventricular valve the authors recommend
inflow or regurgitation that PC CMR
volume can be quantified measurements of peak
indirectly through several velocity be applied
different comparisons of cautiously as there are a
ventricular stroke volume number of factors that may
measured by lead to erroneous velocity
ventriculography and measurements (typically an
semilunar valve flow underestimate) including
measurement by PC CMR difficulty aligning with
[55,88,94,112-115]. complex flow jets, partial
Atrioventricular valve volume effects, insufficient
inflow can also be temporal resolution, signal
quantified directly with PC loss, and misregistration
CMR by prescribing a artifacts.
location perpendicular to As with PC CMR
the inflow direction acquisitions, the post-
[116,117]; to ensure that processing of the image
this plane remains apical to data requires careful
Figure 11 Proper positioning of the imaging plane for main pulmonary the valve throughout the attention to detail. The
artery blood flow measurement. PC CMR velocity and flow
measurements are most accurate when the location of interest is at the isocenter of the scanner during the acquisition. Most MR scanners will
cardiac cycle, ’it should be target vessel must be
slide the patient table so that the center of the imaging plane (yellow circle) is at the scanner
s isocenter (vertical red line). Prescribing the
imaging plane so that the center of the image is at the same level in the positioned
superoinferior at the annular
dimension accurately
as the location of interestidentified
is and a
therefore recommended. level on an endsystolic region of interest should be
drawn on the outer margin motion of the heart and

of the lumen. Automatic blood vessels may impact
border detection tools in the accuracy of flow
most software programs measurements
are fairly accurate; [110,111,118]. 3D PC
however, the contours CMR sequences along with
should be reviewed on specialized post-processing
each image, as the vessel analysis software can
moves and changes size compensate for this motion
during the cardiac cycle. though its use is not yet Figure 12 First-pass perfusion. Patient with anomalous origin of
Abnormal signal, such as widespread [119,125-129]. the left coronary artery from the pulmonary artery who underwent
susceptibility artifact from In all cases, PC CMR data left coronary artery re-implantation and subsequently developed
severe stenosis of the re-implanted coronary artery. First-pass
air in the lungs, should be should be scrutinized to perfusion images at a mid-ventricular level with adenosine stress A)
(
carefully excluded from ensure that they are and at rest (B). Note the extensive sub-endocardial hypoperfusion of
the region of interest. consistent with the known the left ventricle at stress but not at rest which indicates
PC CMR velocity and flow information about the inducible ischemia.
measurement, like all patient’s condition and
quantitative techniques in with the other CMR data
clinical medicine, has from the examination. For
sources of error and instance, the net flow in
limitations, and it is the main pulmonary artery
essential that reporting and ascending aorta should
physicians have a good be approximately equal in
understanding of them. patients without any
These include evidence of a shunt, net
inappropriate setting of the main pulmonary artery
venc, signal loss with flow should equal the sum
complex turbulent flow, of the net branch
partial volume averaging, pulmonary artery flow, and
signal misregistration, and flow in a great vessel
phase offset errors due to should agree well with the
eddy currents or corresponding ventricular
uncorrected concomitant stroke volume from
gradients [70- ventriculography (in the
72,110,120,121]. absence of atrioventricular
Adherence to the valve regurgitation or a
guidelines above will shunt). When data is
minimize these concerns incongruous, one of the
but phase offset errors are known limitations
particularly troublesome as described above can
they are often difficult to usually be identified.
detect and may have a
significant impact on Vasodilator myocardial
accuracy. On some perfusion
scanners it may be Vasodilator perfusion is
advisable to subtract the primarily used to assess
background velocity or patients for inducible
flow measured on a myocardial ischemia
stationary phantom from (Figure 12). It relies
that measured on the
patient’s PC CMR images
[122-124]. Through-plane
on the principle that administration of a coronary artery build an image as is the case for standard cine CMR. A
vasodilator (e.g., adenosine, dipyridamole, and minimum 3 short-axis slices should be prescribed to
regadenoson) will cause a greater increase in the perfusion ensure coverage of all but the apical left ventricular
of myocardium supplied by normal coronary arteries than segment. Additional short-axis or long-axis slices may
myocardium supplied by stenotic coronary arteries. also be useful. At the slower heart rates typically found in
Perfusion is assessed by administering an intravenous adult patients, the cardiac cycle length is long enough to
dose of a GBCA and then rapidly imaging the ventricles allow acquisition of 3–5 slice locations during each beat
at multiple locations to visualize the enhancement pattern (Figure 13). The slice locations are timed to different
during the first transit of the contrast bolus through the phases of the cardiac cycle but each location is acquired
myocardium. The appearance of contrast will be repeatedly at the same phase. At the higher heart rates
attenuated, both in amplitude and rate, in regions of (and shorter cardiac cycle lengths) seen in younger
compromised coronary blood flow. Typically, a perfusion patients, fewer locations can be acquired in one beat.
scan is run both at rest and with vasodilator infusion in Thus, it may be advantageous to spread the imaging
order to distinguish fixed perfusion defects (e.g., infarct) period for the slices over two heartbeats so that a
from inducible ones. As adenosine is the most commonly sufficient number of locations can be acquired–each
used vasodilator medication, the subsequent discussion location every other beat. Spatial resolution should also be
will focus on this agent. increased in smaller patients.
Vasodilator perfusion CMR has been shown to have Detailed instructions on how to perform an adenosine
similar or superior sensitivity and specificity to nuclear perfusion CMR protocol have previously been published
cardiology techniques for detecting significant coronary [143-145]; the basic steps follow here. Prior to
artery stenosis in adult patients [130-133]. Moreover, the undergoing adenosine perfusion CMR, the patient should
results of perfusion CMR have been shown to have be screened for contraindication to adenosine
prognostic value in adults with coronary heart disease administration including second or third degree heart
[134,135]. Reports of adenosine perfusion CMR for the block, sinus node dysfunction, severe asthma or
assessment of coronary artery disease in children and obstructive pulmonary disease, and pregnancy. Caffeine,
adults with CHD have been limited to small studies [136- aminophyline, and nitrate intake should be avoided on the
140]. It is increasingly being used to assess patients with day of the examination as these agents interfere with the
chest pain, anomalous origins of the coronary arteries, and action of adenosine. A specific consent procedure should
after surgeries which involve coronary artery be undertaken to inform the patient/guardian of potential
reimplantation (e.g., the arterial switch or Ross side effects and complications such as dyspnea, flushing,
operations). However, it is important to note that headache, light-headedness, blurred vision, nausea,
adenosine mainly induces myocardial blood flow bronchospasm, heart block, and hypotension. It is
inhomogeneities by vasodilatation and steal effects that do preferable to have two intravenous lines in separate veins
not necessarily represent the pathophysiology of coronary —one for adenosine and one for contrast agent
artery problems found in CHD patients. Furthermore, it is administration—in order to avoid a bolus dose of
unclear whether exercise induced ischemia caused, for adenosine with contrast administration. Monitoring
example, by an anomalous coronary artery coursing equipment should include a continuous ECG recording
between the arterial roots can be detected by adenosine and blood pressure cuff placed on the arm not being used
perfusion CMR. In some settings it may, therefore, be for the adenosine infusion. Equipment and qualified
preferable to use other “stress” agents such as dobutamine personnel for resuscitation must be available.
[141] or another imaging modality in conjunction with Table 5 T1-weighted vasodilator perfusion
exercise stress. saturation-recovery acquisition first followed
Vasodilator perfusion CMR has been performed using a gradient echo for first pass by the rest acquisition
variety of pulse sequences; a detailed review can be found myocardial perfusion with at least 15 minutes
elsewhere [142]. In brief, strong T1 contrast is provided between
Infant/small child the two so as to
by a preparation pulse such as inversion recovery or minimize contamination
saturation recovery (Table 5). This is followed by fast In-plane resolution (mm) of the rest images by
imaging using a gradient echo, gradient echo-planar, or Slice thickness (mm) residual contrast given
SSFP acquisition. Parallel imaging techniques are widely Image acquisition timing 1 R-R during
or 2 R-R the vasodilator
used as a means for accelerating image acquisition. These The consensus group infusion. An adenosine
sequences create images of the heart in one heartbeat authors recommend dose of 0.14 mg/kg/min is
rather than acquiring data from multiple cardiac cycles to performance of the typically used though the
adequacy and safety of infusion is discontinued. (LGE) images [147]. In a true subendocardial
this dose has not been Adenosine should be the absence of LGE, perfusion defect, this dark
thoroughly established in terminated earlier if the homogeneous rim artifact typically lasts
the pediatric age group. patient develops persistent enhancement in all for only a few heartbeats
After the adenosine has or symptomatic heart locations at rest and with and then fades away. The
been infusing for 3 block, significant the vasodilator indicates dark rim artifact is more
minutes, a GBCA (0.05- hypotension, or severe no inducible ischemia. A commonly seen with high
0.1 mmol/ kg) is rapidly respiratory difficulty. An region with transmural heart rates, a more
injected followed by a intravenous dose of LGE and perfusion concentrated contrast
saline flush. In adultsize aminophylline can be used defects at rest and with bolus, and the use of a
patients, rates of 5 ml/s to rapidly counteract the the vasodilator, a balanced SSFP rather than
with 25 ml of flush are effects of adenosine. The socalled “fixed defect”, a gradient echo based
recommended. In same contrast infusion is also interpreted as the perfusion sequences.
children, a minimum rate protocol and pulse absence of inducible
of 3 ml/s through an sequence parameters as ischemia. A region that Late gadolinium
appropriate size used in the adenosine has a perfusion defect enhancement
intravenous line and 10 perfusion segment should with vasodilator but is LGE, also known as
ml flush is suggested. be used later for the rest normal at rest and has no myocardial delayed
The perfusion imaging perfusion imaging. LGE is diagnostic of enhancement, is a
sequence is started Though quantitative ischemia. Perfusion technique that detects
Figure 13 Schematic diagram illustrating data acquisition timing in a first-pass perfusion sequence. Each rectangle represents data
acquisition to form one complete image and the numbers inside them correspond to different slice locations. At a heart rate of 60 bpm A),( the
cardiac cycle length is long enough to allow acquisition of 4 slice locations during each beat. The slice locations are timed to different phases of
the cardiac cycle, but each location is acquired repeatedly at the same phase in subsequent cycles. At a heart rate of 120 bpm B),( the cardiac
cycle length is shorter so only two slice locations can be acquired over each beat; the other 2 locations are acquired the following beat. Note
that the temporal resolution (images per unit time) is the same inA and B.
simultaneously with the analysis of perfusion defects should be focal regions of
contrast agent infusion, imaging [146] is possible reported as either full myocardial fibrosis and
and its scanning duration using commercially thickness or partial infarction. It is based on
should be set to acquire available image analysis thickness (sub- the observation that
approximately 60 software, this process endocardial) with the GBCA have slower
heartbeats. In order to remains time-consuming affected region of the left washout and an increased
minimize respiratory to perform and is subject ventricle described using volume of distribution in
motion artifact, breathing to technical challenges. the standard 17-segment fibrotic and necrotic
should be suspended as Visual analysis by an American Heart myocardium. Thus, such
long as possible during experienced reader is Association model [148]. areas appear brighter than
the image acquisition. If usually sufficient for On some perfusion scans normal myocardium on
breath-holding is not routine clinical practice. there may be a rim of LGE images. Validation
possible, the patient Interpretation is also reduced signal intensity studies have correlated the
should be instructed to guided by review of the in the subendocardium, finding of LGE with the
breathe shallowly. Once cine images of ventricular which can mimic a presence and extent of
the imaging is function and late hypoperfused area myocardial fibrosis
completed, the adenosine gadolinium enhancement [145,149]. Compared to detected by histology in
animal models and in perform a MRA. For gadolinium concentration,

humans [150-152]. LGE imaging, an ECG- an inversion pulse is
Reports in adults have gated 2D incorporated into the pulse
demonstrated its clinical segmented inversion recovery sequence.
gradient The echo pulse time
utility in acute and between the inversion
chronic ischemic heart pulse and image
disease, acquisition, known as the
cardiomyopathies, inversion time (TI),
myocarditis, and should be set to null
ventricular thrombus normal myocardium.
detection. In CHD, LGE Selecting the appropriate
has been described in inversion time may be
patients with repaired facilitated by imaging
tetralogy of Fallot [153- iteratively with different
155], atrial switch Figure 15 Late gadolinium enhancement TIs, orinby the use of a TI-
endocardial
operations for fibroelastosis. Patient with a history scout
of severe or Look-Locker
congenital aortic
valve stenosis who underwent balloon valvuloplasty. Late
transposition of the great sequence. Alternatively, a
gadolinium enhancement images in 4-chamber (A) and short-axis
arteries [156], Fontan Figure 14 Late gadolinium enhancement following
views (B). Note theaextensive phase-sensitive
Fontan subendocardial sequence
late gadolinium
operations for procedure. Patient with tricuspid atresia and normally related great with a standardized
enhancement consistent with endocardial fibroelastosis. TI
arteries who underwent a Fontan procedure. Late gadolinium
functionally single enhancement image in a ventricularsequence is showing
short-axis view typically based on dose, timing, and
ventricles (Figure 14) enhancement and wall thinning of used with consistent
the inferior septum data heart rate may be used
[157], congenitally with a chronic myocardial infarction.acquisition timed to [169]; such sequences
corrected transposition coincide with the cardiac provide consistent contrast
of the great arteries between infarcted/fibrotic
after repair [162], rest period to minimize
[158], aortic valve and normal myocardium
endocardial fibroelastosis motion blurring. Three-
stenosis [159,160], over a wider range of TIs.
(Figure 15) [163,164], and dimensional LGE
pulmonary atresia with Because the gadolinium
fibrous tissue along sequences are available
intact ventricular septum concentration in normal
regions of reconstruction but are less well
[161], anomalous left myocardium decreases
in patients who have had validated [167,168].
coronary artery from the with time, the optimal TI
surgery for CHD [165]. In Breathhold imaging is
pulmonary artery will become longer as
the tetralogy of Fallot, preferred though
time elapses. Accordingly,
atrial switch, and Fontan multiple signal average
imaging, respiratory if the time it takes to
operation cohorts, the
navigator gating, or acquire all of the LGE
presence of LGE has been
images becomes
associated with adverse single-shot imaging can
prolonged, the TI value
ventricular mechanics, be used with free-
may need to be updated to
exercise intolerance, and breathing. Imaging
a longer value. Initial
ventricular arrhythmias planes and slice
thickness should match reports suggested that the
[153,155-158].
inversion time to null
Nevertheless, the those used for cine
myocardial signal is
pathophysiology and imaging of the ventricles
shorter for the RV (in the
prognostic impact of LGE to facilitate comparison.
usual subpulmonary
in patients with CHD has A comprehensive
examination would position) than the LV
not been fully established
include LGE imaging in [170,171]. However, a
[166].
short-axis, LV 2, 3 and more recent study showed
LGE imaging is
that the inversion times to
performed 10–20 minutes 4-chamber, and RV 3-
null myocardium are quite
following injection of 0.1- chamber views.
similar for both ventricles,
0.2 mmol/kg of a GBCA. In order to improve the
and that the apparent
In many imaging image contrast between
difference was related to
protocols, the contrast normal myocardium and
insufficient spatial
dose is initially used to regions of increased
resolution for the thin- Slice thickness (mm) increased and that of usually improved contrast
walled RV [172]. Thus Views per segment abnormal myocardium against the bright
high spatial resolution is decreased (or even enhanced regions.
Number of signal averages
required to properly nulled) leading to gross
Image acquisition timing 3 R-R or 4 R-R
assess for LGE on thin- misinterpretation. If the
Trigger delay Diastole
walled ventricles. TI orissystole
set too long, the Disease-specific protocols
Performing the LGE scanner’s no-trigger relative contrast between This section provides
technique in children interval and double the normal and abnormal suggested disease-specific
requires modifications to data acquisition interval. myocardium will be CMR protocols using the
address smaller-sized In addition, the data reduced and sensitivity modules reviewed above.
ventricles and faster acquisition (shot) duration diminished. For conciseness, the
heart rates (Table 6). In should be decreased by Ghosting artifacts can ventriculography and LGE
order to ensure adequate reducing the views per result from regions that modules are listed once in
spatial resolution, voxel segment (turbo factor) to have a long T1 such as detail at the beginning and
size should be 1.0-1.5 minimize blurring from pericardial fluid or then simply referred to in
mm in-plane with a faster cardiac motion at cerebrospinal fluid [174]. the disease-specific
thickness of 5 mm. The the higher heart rate. Artifact from protocols. Similarly, the
resulting decreased Although image analysis cerebrospinal fluid can abbreviations used in the
signal-to-noise ratio can software can be used to be removed by protocols are listed below
Figure 16 Schematic diagram illustrating data acquisition timing in a late gadolinium enhancement sequence. Each rectangle represents
data acquisition timed to coincide with the cardiac rest period and during which a user-defined number ofk-space lines are filled. Multiple data
acquisitions and thus cardiac cycles are required to fillk-space and produce an image. At a heart rate of 80 bpm(A), data acquisition occurs
every second cardiac cycle in order to allow sufficient time for recovery of longitudinal signal. At a heart rate of 120 bpmB),( the cardiac cycle
length is shorter so the sequence is modified to acquire data every third cycle in order to maintain the same time for the recovery of
longitudinal signal. In addition, the data acquisition duration is shortened to compensate for the briefer cardiac rest period associated with a
faster heart rate.
be compensated for by objectively quantify the Finally, subendocardial, for easy reference. PC
performing two signal extent of enhanced papillary muscle, and CMR acquisitions are with
averages, albeit at the myocardium [173], right ventricular wall throughplane velocity
expense of a longer qualitative visual enhancement may be Ventriculography
acquisition time. At interpretation is typically inconspicuous when
module (see
higher heart rates (>100 performed in clinical adjacent to a relatively
bpm), in order to allow practice. The involved bright blood pool. ventriculography
time for adequate cardiac segments can be Interpreting the LGE section above for
longitudinal signal reported according to the images side-by-side with
recovery between 17-segment left the cine images details) Standard
successive
Table inversion
6 T1-weighted ventricular model [148] addresses this pitfall as imaging
inversion-recovery gradient and a 9-segment right these will provide
echo for late gadolinium ventricular model [155]. information regarding 1. Cine CMR: left
enhancement In addition, it is helpful to wall thickness and ventricular 2-
note the extent of the wall
Infant/small child papillary muscle position chamber and
thickness that is involved [174]. Moreover, 3chamber views,
In-plane resolution (mm) and the pattern of performing LGE imaging
enhancement (e.g., later after contrast agent
right ventricular encompass both

3-chamber view, ventricles planned
4-chamber view in
(Figure 6) a ventricular
2. Cine CMR: stack short-axis or axial
of contiguous orientation
slices 3. LGE imaging in
tocompletely an orthogonal
encompass both plane to confirm
ventricles LGE if present
planned in a
ventricular short-
axis and/or axial
orientation
(Figure 7) Key reporting elements:
Key reporting
location, extent and
elements: left and
right ventricular thickness of LGE
end-diastolic and
end-systolic Coarctation of
volume, stroke
the aorta,
volume, ejection
fraction, and mass;
before or
regional wall
motion after repair
abnormalities
Standard
LGE module
imaging
(see LGE
section
above for
details)
Standard
imaging
1. LGE imaging in
left ventricular 2-
chamber and
3chamber views,
right ventricular 1. Cine CMR: long-
3-chamber view, axis aortic arch
4-chamber view plane (Figure 1)
2. LGE imaging 2. Ventriculography
with a stack of module
contiguous slices
tocompletely
3. CE-MRA or 3D parameters including e, and systemic Standard

SSFP to image and pulmonary
left ventricular mass;
the venous baffles imaging
thoracicvasculatu aortic valve (Figure 4)
re 5. PC CMR: AAo, 1. Ventriculography
morphology,
4. PC CMR: AAo, MPA, branch module
MPA, DAo at stenosis, and PAs, tricuspid 2. Cine CMR: stack
the level of regurgitation andmitral valves of images parallel
thediaphragm Additional case- to the longaxis of
Additional case- specific/comprehensiv the right
D-loop transposition of the
specific/comprehensiv e imaging ventricular
great arteries following an
e imaging outflow tract and
arterial switch operation 1. LGE module
pulmonary valve
1. Spin echo: long- Standard imaging 2. PC CMR: SVC
3. Cine CMR:
axis aortic arch 1. Ventriculography moduledistal to the azygous oblique axial stack
plane (Figure 1) 2. Cine CMR: long-axis of theveinright
and IVCwhen
to image the PAs
2. Cine CMR: systemic venous
ventricularoutflow tract plane
4. CE-MRA or 3D
short-axis aortic 3. Cine CMR: oblique axial obstruction is suspected.
stack to image the PAs
4. CE-MRA or 3D SSFP to image Key the
reporting SSFP to image the
root plane for thoracicvasculatur
valvemorpholog thoracicvasculature elements: location
5. PC CMR: AAo, MPA, branch and PAsseverity of e
y
Additional case-specific/comprehensive
systemic andimaging 5. PC CMR: AAo,
3. PC CMR: MPA, branch PAs
1. 3D SSFP: coronary artery origins and proximal
pulmonary venous
quantification of
course pathway Additional case-
collateral aortic 2. Vasodilator perfusion module obstruction, atrial specific/comprehensive
flowby 3. LGE module
baffle leak, imaging
measuring Key reporting elements: location and severity of
proximal or just ventricular 1. LGE module
AAo, MPA, and branch PA obstruction; branch PA
distal to the parameters,
flow distribution; proximal coronary arteryseverity
patency 2. PC CMR:
coarctation, and and course; aortic root dilation;and mechanism
aortic and of tricuspid and
in the descending left or right
pulmonary regurgitation; ventricular parameters mitral valves
aorta at the level D-loop transposition of ventricular outflow Key reporting elements:
of the diaphragm tract obstruction, location and severity
the great arteries
4. LGE module tricuspid of right ventricular
Key reporting
following an atrial regurgitation, outflow tract and
switch operation Qp/Qs, branch pulmonary artery
elements: location, pulmonary artery obstruction, branch
Standard imaging flow distribution,
dimensions, and pulmonary artery
1. Cine CMR: axial SVC/IVC flow ratio flow distribution,
severity of aortic as an indicator of
stack from the pulmonary
obstruction; arch mid-liver to thetop systemic regurgitation, atrial
of the aortic arch pathway obstruction and ventricular septal
sidedness and
2. Ventriculography Tetralogy of defects, Qp/Qs,
branching order; module ventricular
Fallot
presence of 3. Cine CMR: parameters including
oblique planes to RV volumes and
aneurysm, following
image the SVC ejection fraction
dissection, or andIVC pathways complete
in long-axis Se
collateral vessels to
4. CE-MRA or 3D repair
the DAo; c
SSFP to image the
thoracicvasculatur
ventricular u
n a r anoblique sagittal
plane
d l d perpendicular to
the atrial septum
u to completely
encompass the
m d i atrial septum
(Figure 17)
e 3. Ventriculography
m
module
a f
a 4. PC CMR: AAo,
MPA
t e
g Additional case-
r c specific/comprehensive
i imaging
i t
1. PC CMR: 1–3
n
contiguous slices
a
g positioned parallel
to the atrial septal
l S
1. Cine CMR: stack plane to obtain an
t of contiguous en face view of
thin slices in a the defect (Figure
s a 4chamber plane 17), and with
to completely through-plane
e
n encompass the velocity encoding
atrial septum 2. PC CMR: stack of
p contiguous thin
d 2. Cine CMR: stack
of contiguous slices in a
t
a thin slices in 4chamber plane
Figure 17 Secundum atrial septal defect (ASD) imaging protocol. A.Axial steady-state free precession (SSFP) image in a 10-year-old with a
large secundum ASD (white arrow).B. The axial SSFP image is used to plan a stack of oblique sagittal SSFP images to visualize the ASD and the
superior and inferior defect margins. C. The axial and oblique sagittal images are used together to plan a stack of phase-contrast (PC) cine images
to visualize the ASD flowen face. This provides insight into the oval shape of the defect and may demonstrate additional ASDs.
and/or in an mid-abdomen s c
oblique sagittal tothe top of the
plane aortic arch t
perpendicular to 2. Cine CMR:
the atrial septum v
oblique plane to
to completely image the
e S
encompass the anomalousvein(s)
atrial septum, in long-axis n t
and with in-plane 3. Ventriculography
velocity module o a
encoding in the 4. CE-MRA or 3D
direction of atrial SSFP to image the s
n
septal defect thoracicvasculatur
flow e (Figure 3) u
d
3. CE-MRA or 3D 5. PC CMR: AAo,
SSFP to image MPA, branch PAs s
a
the Additional case-
thoracicvasculatu specific/comprehensive r
re imaging s
Key reporting d
1. PC CMR:
elements: number anomalous vein e
and location of the Key reporting elements:
p
defects, defect rim number, location, i
measurements, and drainage of the t
ventricular pulmonary veins; m
a
parameters, Qp/Qs ventricular
a
parameters; Qp/Qs, l
Partially anomalous
branch pulmonary g
pulmonary venous
artery flow
connection, before or distribution i
d
Figure 18 Sinus venosus septal defect imaging protocol. A.Axial steady-state free precession (SSFP) cine image in a 22-year-old with a large
sinus venosus septal defect (white arrow).B. The axial SSFP image is used to plan a stack of oblique sagittal SSFP cine images to visualize the
defect in an orthogonal view and assess its superoinferior dimension.
after repair Standard Si e n
imaging n f g
1. Cine CMR: axial
stack from the u e
1. Cine CMR: axial 2. Ventriculography module with multiplecontiguous

parameters, 3. Spin echo: axial
stack from the slices in the right ventricular 3-chamber
presenceview
of anand in
atrial plane to image the
mid-liver to the 4-chamber view septal defect, Qp/Qs branch PAsand
thetop of the 3. PC CMR: AAo, MPA, tricuspid based valve,mitral
on flowvalve aortopulmonary
aortic arch measurements shunt
Additional case-specific/comprehensive imaging
2. Cine CMR: 4. PC CMR: SVC,
1. Cine CMR: contiguous stack oriented parallel tothe
oblique sagittal Functional single ventricle IVC, tricuspid and
functional tricuspid valve plane to visualize the
plane following a stage 1 or 2 mitral valves,DAo
valve en face (Figure 19)
perpendicularto palliation at the level of the
2. CE-MRA or 3D SSFP to image the
the plane of the diaphragm,
thoracicvasculature
defect (Figure Standard imaging pulmonary veins
Key reporting elements: description of tricuspid
18) valve morphology, tricuspid regurgitation and axial
1. Cine CMR: Key reporting elements:
3. Ventriculograph stenosis, pulmonary stenosis, ventricular
stack from the shunt, branch PA,
y module Ebstein anomaly of the mid-liver to pulmonary vein, and
4. CE-MRA or 3D tricuspid valve thetop of the aortic arch
SSFP to image aortic arch obstruction;
the 2. Ventriculograph ventricular
thoracicvasculatu y module parameters; valve
re 3. CE-MRA or 3D regurgitation;
5. PC CMR: AAo, SSFP to image aortopulmonary
MPA the collaterals; venous
Key reporting thoracicvasculatu collaterals
elements: location re and surgical
shunts Functional single ventricle
and size of the following a Fontan
Figure 19 Ebstein anomaly imaging protocol. A.Right ventricular 3-chamber view (RV 3C) steady-state free precession cine image in a 22-year
-old with Ebstein anomaly. In this example, the functional tricuspid valve plane is displaced and inflow is directed into the right ventricular
outflow tract (white arrow). B. The RV 3C image is used to plan a stack of cine images to visualize the displaced tricuspid valve orifice
en face for
anatomic assessment or flow quantification.
defect, drainage of 4. PC CMR: AAo, operation
the right pulmonary native MPA,

branch PAs Standard imaging
veins, ventricular Additional case- 1. Cine CMR: axial
parameters, Qp/Qs specific/comprehensiv stack from the
e imaging mid-liver to thetop
based on flow
1. LGE module of the aortic arch
measurements 2. Cine CMR: long- 2. Cine CMR:
axis aortic arch coronal or oblique
Standard imaging plane stack to image
1. Cine CMR: axial stack from the diaphragm to thetop theFontan baffle
of the aortic arch in long-axis
3. Ventriculograph Left ventricle; IVC: Inferior vena authors thank Emily Harris for S, Stern H. Routine clinical
cava; MPA: Main pulmonary her assistance with the cardiovascular magnetic
y module resonance in paediatric and adult
artery; MR: Magnetic resonance; preparation of the figures.
4. CE-MRA or 3D MRA: Magnetic resonance congenital heart disease: patients,
SSFP to image angiography; NSF: Nephrogenic Author details protocols, questions asked and
systemic fibrosis; PAs: Pulmonary 1 Department of contributions made. J Cardiovasc
the Pediatric Cardiology Magn Reson. 2008; 10:46.
arteries; PC CMR: Velocity-
thoracicvasculatu encoded phase-contrast cine CMR; and Congenital Heart 4. Bailliard F, Hughes ML, Taylor AM.
re Qp/ Qs: Pulmonary-to-systemic Disease, Introduction to cardiac imaging in
Deutsches Herzzentrum infants and children: techniques,
5. PC CMR: AAo, flow ratio; RV: Right ventricle;
München (German Heart Center
RVOT: Right ventricular outflow potential, and role in the imaging
native MPA, Munich) of the work-up of various cardiac
tract; SA: Short-axis; SSFP: Steady-
branch state free precession; SVC:
Technical University Munich, malformations and other pediatric
Munich, Germany. 2Department heart conditions. Eur J Radiol.
PAs,SVC, IVC Superior vena cava; TE: Echo time;
of Diagnostic 2008; 68:191–98.
Additional case- TFE: Turbo field echo; TI: Inversion
Imaging, Children’s Hospital & 5. Marcotte F, Poirier N, Pressacco J,
time; TR: Repetition time; venc:
specific/comprehensiv Velocity encoding range. Research Center Oakland, Paquet E, Mercier LA, Dore A,
e imaging Oakland, California, USA.
3Department of Pediatric
Ibrahim R, Khairy P. Evaluation of
adult congenital heart disease by
1. Cine CMR: long- Competing interests Cardiology, Evelina Children’s cardiac magnetic resonance
The authors declare that they have Hospital/Guy’s and St. Thomas’ imaging. Congenit Heart Dis. 2009;
axis aortic arch no competing interests. Hospital NHS Foundation Trust; 4:216–30.
plane Division of Imaging Sciences & 6. Prakash A, Powell AJ, Geva T.
2. LGE module Authors’ contributions
Biomedical Engineering, King’s Multimodality noninvasive
College London, London, UK.
3. PC CMR: SF drafted and conceived of the 4The Herma
imaging for assessment of
congenital heart disease. Circ
tricuspid and review, designed the initial
Heart Center, Children’s Hospital Cardiovasc Imaging. 2010; 3:112–
structure of the manuscript,
mitral coordinated author
of Wisconsin, Medical College of 25.
Wisconsin, Milwaukee, WI, USA.
valves,pulmonar communication, and revised the 5Centre for Cardiovascular
7. Kilner PJ, Geva T, Kaemmerer H,
Trindade PT, Schwitter J, Webb
y veins authors’ contributions to produce Imaging, UCL Institute of GD. Recommendations for
the draft version of the
Key reporting Cardiovascular Science, & Great cardiovascular magnetic
manuscript; TC drafted the Ormond Street Hospital for resonance in adults with
elements: Fontan ventriculography section; GFG Children, London, congenital heart disease from the
drafted the contrast-enhanced
pathway, SVC, UK. 6Division of Cardiology, respective working groups of the
MRA and 3D SSFP sections; MMS University Children’s Hospital European Society of Cardiology.
IVC, branch PA, provided extensive editorial Zurich, Zurich, Eur Heart J. 2010; 31:794–805.
assistance; AMT drafted of the Switzerland. 7Department of
pulmonary vein, 8. Zikria JF, Machnicki S, Rhim E,
vasodilator perfusion section; Diagnostic Imaging and Division Bhatti T, Graham RE. MRI of
and aortic arch ERVB drafted the sedation section; of Cardiology, patients with cardiac pacemakers:
SJY drafted the blood velocity and Department of Paediatrics,
obstruction; Fontan flow measurement section; AJP Hospital for Sick Children,
a review of the medical literature.
AJR Am J Roentgenol. 2011;
baffle defects; drafted the LGE and patient University of Toronto, 196:390–401.
preparation sections, and revised Toronto, Ontario, Canada.
ventricular and synthesized all of the authors’ 8Department of Cardiology,
9. Levine GN, Gomes AS, Arai AE,
Bluemke DA, Flamm SD, Kanal E,
parameters; valve contributions to produce the final Boston Children’s Hospital, and Manning WJ, Martin ET, Smith JM,
version of the manuscript. All the Department of Pediatrics,
regurgitation; Harvard Medical School, Boston,
Wilke N, Shellock FS. Safety of
authors contributed substantially magnetic resonance imaging in
aortopulmonary to the design of the review, MA, USA. patients with cardiovascular
collaterals critically revised the manuscript devices: an American Heart
Received: 9 April 2013 Accepted:
for important intellectual content, Association scientific statement
Abbreviations 8 May 2013
and read and approved the final from the Committee on
2C: 2-chamber; 2D: 2- Published: 13 June 2013
manuscript. Diagnostic and Interventional
dimensional; 3C: 3-chamber; 4C: Cardiac Catheterization,
4-chamber; 3D: 3dimensional; References Council on
Acknowledgements
AAo: Ascending aorta; ASD: 1. Weber OM, Higgins CB. MR Clinical Cardiology, and the
The authors thank Mark A. Fogel,
Atrial septal defect; CEMRA: Evaluation of Cardiovascular Council on Cardiovascular
MD, from The Children’s Hospital
Contrast-enhanced magnetic Physiology in Congenital Heart Radiology and
of Philadelphia, PA, USA, for
resonance angiography; CHD: Disease: Flow and Function. J Intervention: endorsed by the
critically reviewing the final
Congenital heart disease; CMR: Cardiovasc Magn Reson. 2006; American College of
version of the manuscript. All
Cardiovascular magnetic 8:607–17. Cardiology
authors were supported by their
resonance; DAo: Descending 2. Crean A. Cardiovascular MR and Foundation, the North
institutions, respectively. The
aorta; ECG: Electrocardiogram; CT in congenital heart disease. American Society for Cardiac
Society of Cardiovascular Magnetic Heart. 2007; 93:1637–47.
GBCA: Gadolinium-based Imaging, and the
Resonance funded a large portion 3. Fratz S, Hess J, Schuhbaeck A,
contrast agents; LGE: Late Society for Cardiovascular
of the telephone communications Buchner C, Hendrich E, Martinoff
gadolinium enhancement; LV: Magnetic Resonance.
during the drafting process. The
Circulation. 2007; 19. Runge VM. Safety of approved MR children. J Magn Reson Imaging. Juvenile and adult congenital
116:2878–91. contrast media for intravenous 2000; 12:1034–42. heart disease: time-resolved 3D
10. Shellock FG, Spinazzi A. MRI injection. J Magn Reson Imaging. 30. Prasad SK, Soukias N, Hornung T, contrast-enhanced MR
Safety Update 2008: Part 2, 2000; 12:205–13. Khan M, Pennell DJ, Gatzoulis angiography. Radiology. 2007;
Screening Patients for MRI. Am J 20. Prince MR, Zhang H, Zou Z, Staron MA, Mohiaddin RH. Role of 244:399–410.
Roentgenol. 2008; 191:1140–49. RB, Brill PW. Incidence of magnetic resonance angiography 38. Goo HW, Yang DH, Park IS, Ko JK,
11. Fogel MA, Pawlowski TW, Harris immediate gadolinium contrast in the diagnosis of major Kim YH, Seo DM, Yun TJ, Park JJ.
MA, Whitehead KK, Keller MS, media reactions. AJR Am J aortopulmonary collateral Time-resolved three-dimensional
Wilson J, Tipton D, Harris C. Roentgenol. 2011; 196:W138–143. arteries and partial anomalous contrast-enhanced magnetic
Comparison and usefulness of 21. Dillman JR, Ellis JH, Cohan RH, pulmonary venous drainage. resonance angiography in patients
cardiac magnetic resonance Strouse PJ, Jan SC. Frequency and Circulation. 2004; 109:207–14. who have undergone a Fontan
versus computed tomography in severity of acute allergic-like 31. Ferrari VA, Scott CH, Holland GA, operation or bidirectional
infants six months of age or reactions to gadolinium-containing Axel L, Sutton MS. Ultrafast cavopulmonary connection: initial
younger with aortic arch i.v. contrast media in children and threedimensional contrast- experience. J Magn Reson
anomalies without deep adults. AJR Am J Roentgenol. enhanced magnetic resonance Imaging. 2007; 25:727–36.
sedation or anesthesia. Am J 2007; 189:1533–38. angiography and imaging in the 39. Muthupillai R, Vick GW 3rd,
Cardiol. 2011; 108:120–25. 22. ACR Manual on Contrast Media diagnosis of partial anomalous Flamm SD, Chung T. Time-resolved
12. Windram J, Grosse-Wortmann L, Version 9. 2013. pulmonary venous drainage. J contrastenhanced magnetic
Shariat M, Greer ML, Crawford http://www.acr.org/QualitySafety Am Coll Cardiol. 2001; 37:1120– resonance angiography in
MW, Yoo SJ. Cardiovascular MRI /Resources/Contrast-Manual. 28. pediatric patients using sensitivity
without sedation or general 23. Leiner T, Kucharczyk W. NSF 32. Valsangiacomo Buchel ER, encoding. J Magn Reson Imaging.
anesthesia using a feed-and prevention in clinical practice: DiBernardo S, Bauersfeld U, 2003; 17:559–64.
-sleep technique in neonates summary of recommendations Berger F. Contrast-enhanced 40. Sorensen TS, Korperich H, Greil
and infants. Pediatr Radiol. and guidelines in the United magnetic resonance angiography GF, Eichhorn J, Barth P, Meyer H,
2011; 42:183–87. States, Canada, and Europe. J of the great arteries in patients Pedersen EM, Beerbaum P.
13. Odegard KC, DiNardo JA, Tsai- Magn Reson Imaging. 2009; with congenital heart disease: an Operator-independent isotropic
Goodman B, Powell AJ, Geva T, 30:1357–63. accurate tool for planning three-dimensional magnetic
Laussen PC. 24. Penfield JG. Nephrogenic systemic catheter-guided interventions. resonance imaging for
Anaesthesia considerations fibrosis and the use of Int J Cardiovasc Imaging. 2005; morphology in congenital heart
for cardiac MRI in infants gadoliniumbased contrast agents. 21:313–22. disease: a validation study.
and small children. Paediatr Pediatr Nephrol. 2008; 23:2121– 33. Geva T, Greil GF, Marshall AC, Circulation. 2004; 110:163–69.
Anaesth. 2004; 14:471–76. 29. Landzberg M, Powell AJ. 41. Razavi RS, Hill DL, Muthurangu V,
14. Dorfman AL, Odegard KC, Powell 25. Zou Z, Zhang HL, Roditi GH, Leiner Gadolinium-enhanced 3- Miquel ME, Taylor AM, Kozerke S,
AJ, Laussen PC, Geva T. Risk T, Kucharczyk W, Prince MR. dimensional magnetic resonance Baker EJ. Three-dimensional
factors for adverse events during angiography of pulmonary blood magnetic resonance imaging of
Nephrogenic systemic
cardiovascular magnetic supply in patients with complex congenital cardiac anomalies.
fibrosis: review of 370 biopsy-
resonance in congenital heart pulmonary stenosis or atresia: Cardiol Young. 2003; 13:461–65.
confirmed cases. JACC
disease. J Cardiovasc Magn comparison with x-ray 42. Fenchel M, Greil GF, Martirosian P,
Cardiovasc Imaging. 2011;
Reson. 2007; 9:793–98. angiography. Circulation. 2002; Kramer U, Schick F, Claussen CD,
4:1206–16.
15. Osborn I. Magnetic resonance 106:473–78. Sieverding L, Miller S. Three-
26. Bernstein MA, Huston J 3rd, Ward
imaging anesthesia: new 34. Greil GF, Powell AJ, Gildein HP, dimensional morphological
HA. Imaging artifacts at 3.0T. J
challenges and techniques. Curr Geva T. Gadolinium-enhanced magnetic resonance imaging in
Magn Reson Imaging. 2006;
Opin Anaesthesiol. 2002; threedimensional magnetic infants and children with
24:735–46.
15:443–48. resonance angiography of congenital heart disease. Pediatr
27. Mueller A, Kouwenhoven M,
16. Fogel MA, Weinberg PM, Parave pulmonary and systemic venous Radiol. 2006; 36:1265–72.
Naehle CP, Gieseke J, Strach K,
E, Harris C, Montenegro L, Harris anomalies. J Am Coll Cardiol. 43. Beerbaum P, Sarikouch S, Laser KT,
Willinek WA, Schild HH, Thomas D.
MA, Concepcion M. Deep 2002; 39:335–41. Greil G, Burchert W, Korperich H.
Dual-source radiofrequency
sedation for cardiac magnetic 35. Saleh RS, Patel S, Lee MH, Coronary anomalies assessed by
transmission with patient-adaptive
resonance imaging: a Boechat MI, Ratib O, Saraiva CR, whole-heart isotropic 3D magnetic
local radiofrequency shimming for
comparison with cardiac Finn JP. Contrast-enhanced MR resonance imaging for cardiac
3.0-T cardiac MR imaging: initial
anesthesia. J Pediatr. 2008; angiography of the chest and morphology in congenital heart
experience. Radiology. 2012;
abdomen with use of controlled disease. J Magn Reson Imaging.
152:534–39. 539 e531. 263:77–85.
apnea in children. Radiology. 2009; 29:320–27.
17. Didier D, Ratib O, Beghetti M, 28. Garg R, Powell AJ, Sena L, Marshall
2007; 243:837–46. 44. Rajiah P, Setser RM, Desai MY,
Oberhaensli I, Friedli B. AC, Geva T. Effects of metallic
36. Mohrs OK, Petersen SE, Flamm SD, Arruda JL. Utility of
Morphologic and functional implants on magnetic resonance
Voigtlaender T, Peters J, Nowak freebreathing, whole-heart, three-
evaluation of congenital heart imaging evaluation of Fontan
B, Heinemann MK, Kauczor HU. dimensional magnetic resonance
disease by magnetic resonance palliation. Am J Cardiol. 2005;
Time-resolved contrast- imaging in the assessment of
imaging. J Magn Reson Imaging. 95:688–91.
enhanced MR angiography of coronary anatomy for congenital
1999; 10:639–55. 29. Masui T, Katayama M, Kobayashi
the thorax in adults with heart disease. Pediatr Cardiol.
18. Sarikouch S, Schaeffler R, S, Ito T, Seguchi M, Koide M,
congenital heart disease. AJR Am 2011; 32:418–25.
Korperich H, Dongas A, Haas NA, Nozaki A, Sakahara H. Gadolinium-
J Roentgenol. 2006; 187:1107– 45. Su JT, Chung T, Muthupillai R,
Beerbaum P. Cardiovascular enhanced MR angiography in the
14. Pignatelli RH, Kung GC, Diaz LK,
magnetic resonance imaging for evaluation of congenital
37. Fenchel M, Saleh R, Dinh H, Lee Vick GW 3rd, Kovalchin JP.
intensive care infants: safe and cardiovascular disease pre- and
effective? Pediatr Cardiol. 2009; MH, Nael K, Krishnam M, Ruehm Usefulness of real-time navigator
postoperative states in infants and
30:146–52. SG, Miller S, Child J, Finn JP. magnetic resonance imaging for
evaluating coronary artery contrast MRI flow. Circ ventricular measurements and Interstudy variability in cardiac
origins in pediatric patients. Am Cardiovasc Imaging. 2010; measurement reproducibility at magnetic resonance imaging
J Cardiol. 2005; 95:679–82. 3:65–76. cardiovascular MR imaging. measurements of ventricular
46. Tangcharoen T, Bell A, Hegde S, 53. Buechel EV, Kaiser T, Jackson C, Radiology. 2005; 236:57–64. volume, mass, and ejection
Hussain T, Beerbaum P, Schmitz A, Kellenberger CJ. 61. Janik M, Cham MD, Ross MI, fraction in repaired tetralogy of
Schaeffter T, Razavi R, Botnar Normal rightand left ventricular Wang Y, Codella N, Min JK, Fallot: A prospective observational
RM, Greil GF. Detection of volumes and myocardial mass in Prince MR, Manoushagian S, study. J Magn Reson Imaging.
coronary artery anomalies in children measured by steady state Okin PM, Devereux RB, Weinsaft 2013. doi:10.1002/jmri.24050.
infants and young children with free precession cardiovascular JW. Effects of papillary muscles 69. Buechel ER, Dave HH, Kellenberger
congenital heart disease by magnetic resonance. J Cardiovasc and trabeculae on left CJ, Dodge-Khatami A, Pretre R,
using MR imaging. Magn Reson. 2009; 11:19. ventricular quantification: Berger F, Bauersfeld U.
Radiology. 2011; 259:240– 54. Alfakih K, Plein S, Bloomer T, Jones increased impact of Remodelling of the right ventricle
47. T, Ridgway J, Sivananthan M. methodological variability in after early pulmonary valve
47. Plathow C, Ley S, Zaporozhan J, Comparison of right ventricular patients with left ventricular replacement in children with
Schobinger M, Gruenig E, volume measurements between hypertrophy. J Hypertens. 2008; repaired tetralogy of Fallot:
Puderbach M, Eichinger M, axial and short axis orientation 26:1677–85. assessment by cardiovascular
Meinzer HP, Zuna I, Kauczor HU. using steady-state free precession 62. Alfakih K, Plein S, Thiele H, Jones magnetic resonance. Eur Heart J.
Assessment of reproducibility magnetic resonance imaging. J T, Ridgway JP, Sivananthan MU. 2005; 26:2721–27.
and stability of different breath- Magn Reson Imaging. 2003; Normal human left and right 70. Powell AJ, Geva T. Blood Flow
hold maneuvers by dynamic 18:25–32. ventricular dimensions for MRI Measurement by Magnetic
MRI: comparison between 55. Fratz S, Janello C, Muller D, as assessed by turbo gradient Resonance Imaging in Congenital
healthy adults and patients with Seligmann M, Meierhofer C, echo and steady-state free Heart Disease. Pediatr Cardiol.
pulmonary hypertension. Eur Schuster T, Schreiber C, Martinoff precession imaging sequences. J 2000; 21:47–58.
Radiol. 2006; 16:173–79. S, Hess J, Kuhn A, et al. The Magn Reson Imaging. 2003; 71. Lotz J, Meier C, Leppert A,
48. Holland AE, Goldfarb JW, functional right ventricle and 17:323–29. Galanski M. Cardiovascular flow
Edelman RR. Diaphragmatic and tricuspid regurgitation in Ebstein’s 63. Hudsmith LE, Petersen SE, measurement with phase-contrast
cardiac motion during anomaly. Int J Cardiol. 2013; Francis JM, Robson MD, MR imaging: basic facts and
suspended breathing: 167:258–61. Neubauer S. Normal human left implementation. RadioGraphics.
preliminary experience and 56. Mooij CF, de Wit CJ, Graham DA, and right ventricular and left 2002; 22:651–71.
implications for breath-hold MR Powell AJ, Geva T. Reproducibility atrial dimensions using steady 72. Gatehouse PD, Keegan J, Crowe
imaging. Radiology. 1998; of MRI measurements of right state free precession magnetic LA, Masood S, Mohiaddin RH,
209:483–89. ventricular size and function in resonance imaging. J Cardiovasc Kreitner KF, Firmin DN.
49. Foo TK, Bernstein MA, Aisen AM, patients with normal and dilated Magn Reson. 2005; 7:775–82. Applications of phase-contrast
Hernandez RJ, Collick BD, ventricles. J Magn Reson Imaging. 64. Sluysmans T, Colan SD. flow and velocity imaging in
Bernstein T. Improved ejection 2008; 28:67–73. Theoretical and empirical cardiovascular MRI. Eur Radiol.
fraction and flow velocity 57. Kirschbaum SW, Baks T, derivation of cardiovascular 2005; 15:2172–84.
estimates with use of view Gronenschild EH, Aben JP, allometric relationships in 73. Chai P, Mohiaddin R. How we
sharing and uniform repetition Weustink AC, Wielopolski PA, children. J Appl Physiol. 2005; perform cardiovascular magnetic
time excitation with fast cardiac Krestin GP, de Feyter PJ, van 99:445–57. resonance flow assessment using
techniques. Radiology. 1995; Geuns RJ. Addition of the long-axis 65. Dewey FE, Rosenthal D, Murphy phase-contrast velocity mapping. J
195:471–78. information to short-axis contours DJ Jr, Froelicher VF, Ashley EA. Cardiovasc Magn Reson. 2005;
50. Robbers-Visser D, Boersma E, reduces interstudy variability of Does size matter? Clinical 7:705–16.
Helbing WA. Normal left-ventricular analysis in cardiac applications of scaling cardiac 74. Hundley WG, Li HF, Hillis LD,
biventricular function, volumes, magnetic resonance studies. size and function for body size. Meshack BM, Lange RA, Willard JE,
and mass in children aged 8 to Invest Radiol. 2008; 43:1–6. Circulation. 2008; 117:2279–87. Landau C, Peshock RM.
17 years. J Magn Reson Imaging. 58. Winter MM, Bernink FJ, Groenink 66. Luijnenburg SE, Robbers-Visser Quantitation of cardiac output
2009; 29:552–59. M, Bouma BJ, van Dijk AP, Helbing D, Moelker A, Vliegen HW, with velocity-encoded,
51. Fratz S, Schuhbaeck A, Buchner WA, Tijssen JG, Mulder BJ. Mulder BJ, Helbing WA. Intra- phasedifference magnetic
C, Busch R, Meierhofer C, Evaluating the systemic right observer and interobserver resonance imaging. Am J Cardiol.
Martinoff S, Hess J, Stern H. ventricle by CMR: the importance variability of biventricular 1995; 75:1250–55.
Comparison of accuracy of axial of consistent and reproducible function, volumes and mass in 75. Fratz S, Hager A, Busch R,
slices versus short-axis slices for delineation of the cavity. J patients with congenital heart Kaemmerer H, Schwaiger M,
measuring ventricular volumes Cardiovasc Magn Reson. 2008; disease measured by CMR Lange R, Hess J, Stern HC. Patients
by cardiac magnetic resonance 10:40. imaging. Int J Cardiovasc after atrial switch operation for
in patients with corrected 59. Sievers B, Kirchberg S, Bakan A, Imaging. 2010; 26:57–64. transposition of the great arteries
tetralogy of fallot. Am J Cardiol. Franken U, Trappe HJ. Impact of 67. Beerbaum P, Barth P, Kropf S, can not increase stroke volume
2009; 103:1764–69. papillary muscles in ventricular Sarikouch S, Kelter-Kloepping A, under dobutamine stress as
52. Sarikouch S, Peters B, Gutberlet volume and ejection fraction Franke D, Gutberlet M, Kuehne opposed to patients with
M, Leismann B, Kelter-Kloepping assessment by cardiovascular T. Cardiac function by MRI in congenitally corrected
A, Koerperich H, Kuehne T, magnetic resonance. J Cardiovasc congenital heart disease: impact transposition. Circ J. 2008;
Beerbaum P. Sex-specific Magn Reson. 2004; 6:9–16. of consensus training on 72:1130–35.
pediatric percentiles for 60. Papavassiliu T, Kuhl HP, Schroder interinstitutional variance. J 76. Hundley WG, Li HF, Lange RA,
ventricular size and mass as M, Suselbeck T, Bondarenko O, Magn Reson Imaging. 2009; Pfeifer DP, Meshack BM, Willard
reference values for cardiac MRI: Bohm CK, Beek A, Hofman MM, 30:956–66. JE, Landau C, Willett D, Hillis LD,
assessment by steady-state van Rossum AC. Effect of 68. Blalock SE, Banka P, Geva T, Peshock RM. Assessment of left-
free-precession and phase- endocardial trabeculae on left Powell AJ, Zhou J, Prakash A. to-right intracardiac shunting by
velocity-encoded, phase- perfusion scintigraphy. Pediatr mass studied with MR imaging in resonance imaging. J Am Coll
difference magnetic resonance Radiol. 2005; 35:295–301. children with pulmonary Cardiol. 1996; 28:1818–26.
imaging. A comparison with 84. Caputo GR, Kondo C, Masui T, regurgitation after repair for 101. Mohiaddin RH, Kilner PJ, Rees S,
oximetric and indicator dilution Geraci SJ, Foster E, O’Sullivan MM, tetralogy of Fallot. Radiology. Longmore DB. Magnetic
techniques. Circulation. 1995; Higgins CB. Right and left lung 1996; 201:135–40. resonance volume flow and jet
91:2955–60. perfusion: in vitro and in vivo 93. Dulce MC, Mostbeck GH, velocity mapping in aortic
77. Brenner LD, Caputo GR, validation with obliqueangle, O’Sullivan M, Cheitlin M, Caputo coarctation. J Am Coll Cardiol.
Mostbeck G, Steiman D, Dulce velocity-encoded cine MR imaging. GR, Higgins CB. Severity of aortic 1993; 22:1515–21.
M, Cheitlin MD, O’Sullivan M, Radiology. 1991; 180:693–98. regurgitation: interstudy 102. Sakuma H, Kawada N, Kubo H,
Higgins CB. Quantification of left 85. Sridharan S, Derrick G, Deanfield J, reproducibility of measurements Nishide Y, Takano K, Kato N,
to right atrial shunts with Taylor AM. Assessment of with velocity-encoded cine MR Takeda K. Effect of breath holding
velocity-encoded cine nuclear differential branch pulmonary imaging. Radiology. 1992; on blood flow measurement using
magnetic resonance imaging. J blood flow: a comparative study of 185:235–40. fast velocity encoded cine MRI.
Am Coll Cardiol. 1992; 20:1246– phase contrast magnetic 94. Mahle WT, Parks WJ, Fyfe DA, Magn Reson Med. 2001; 45:346–
50. resonance imaging and Sallee D. Tricuspid regurgitation 48.
78. Beerbaum P, Korperich H, Barth radionuclide lung perfusion in patients with repaired 103. Johansson B, Babu-Narayan SV,
P, Esdorn H, Gieseke J, Meyer H. imaging. Heart. 2006; 92:963–68. Tetralogy of Fallot and its Kilner PJ. The effects of breath-
Noninvasive quantification 86. Holmqvist C, Oskarsson G, relation to right ventricular holding on pulmonary
of left-to-right shunt in Stahlberg F, Thilen U, Bjorkhem G, dilatation. Am J Cardiol. 2003; regurgitation measured by
pediatric patients: phase- Laurin S. Functional evaluation of 92:643–45. cardiovascular magnetic
contrast cine magnetic extracardiac ventriculopulmonary 95. Whitehead KK, Gillespie MJ, resonance velocity mapping. J
resonance imaging conduits and of the right ventricle Harris MA, Fogel MA, Rome JJ. Cardiovasc Magn Reson. 2009;
compared with invasive with magnetic resonance imaging Noninvasive quantification of 11:1.
oximetry. Circulation. 2001; and velocity mapping. Am J systemic-to-pulmonary collateral 104. Ley S, Fink C, Puderbach M,
103:2476–82. Cardiol. 1999; 83:926–32. flow: a major source of Zaporozhan J, Plathow C, Eichinger
79. Powell AJ, Maier SE, Chung T, 87. Rebergen SA, van der Wall EE, inefficiency in patients with M, Hosch
Geva T. Phase-velocity cine Doornbos J, de Roos A. Magnetic superior cavopulmonary W, Kreitner KF, Kauczor HU. MRI
magnetic resonance imaging resonance measurement of connections. Circ Cardiovasc Measurement of the
measurement of pulsatile blood velocity and flow: technique, Imaging. 2009; 2:405–11. hemodynamics of
flow in children and young validation, and cardiovascular 96. Grosse-Wortmann L, Al-Otay A, the pulmonary and systemic
adults: in vitro and in vivo applications. Am Heart J. 1993; Yoo SJ. Aortopulmonary arterial circulation: influence
validation. Pediatr Cardiol. 2000; 126:1439–56. collaterals after bidirectional of breathing maneuvers. AJR
21:104–10. 88. Kon MW, Myerson SG, Moat NE, cavopulmonary connection or Am J Roentgenol. 2006;
80. Beerbaum P, Korperich H, Pennell DJ. Quantification of Fontan completion: 187:439–44.
Gieseke J, Barth P, Peuster M, regurgitant fraction in mitral quantification with MRI. 105. Prakash A, Garg R, Marcus EN,
Meyer H. Rapid left-to-right regurgitation by cardiovascular Circ Cardiovasc Imaging. Reynolds G, Geva T, Powell AJ.
shunt quantification in children magnetic resonance: 2009; 2:219–25. Faster flow quantification using
by phase-contrast magnetic comparison of techniques. J 97. Prakash A, Rathod RH, Powell AJ, sensitivity encoding for velocity-
resonance imaging combined Heart Valve Dis. 2004; McElhinney DB, Banka P, Geva T. encoded cine magnetic resonance
with sensitivity encoding 13:600–07. Relation of systemic-to- imaging: in vitro and in vivo
(SENSE). Circulation. 2003; 89. Kutty S, Whitehead KK, Natarajan pulmonary artery collateral flow validation. J Magn Reson Imaging.
108:1355–61. S, Harris MA, Wernovsky G, Fogel in single ventricle physiology to 2006; 24:676–82.
81. Powell AJ, Tsai-Goodman B, MA. Qualitative echocardiographic palliative stage and clinical 106. Thunberg P, Karlsson M, Wigstrom
Prakash A, Greil GF, Geva T. assessment of aortic valve status. Am J Cardiol. 2012; L. Accuracy and reproducibility in
Comparison between phase- regurgitation with quantitative 109:1038–45. phase contrast imaging using
velocity cine magnetic cardiac magnetic resonance: a 98. Rupprecht T, Nitz W, Wagner M, SENSE. Magn Reson Med. 2003;
resonance imaging and invasive comparative study. Pediatr Kreissler P, Rascher W, Hofbeck 50:1061–68.
oximetry for quantification of Cardiol. 2009; 30:971–77. M. Determination of the 107. Tang C, Blatter DD, Parker DL.
atrial shunts. Am J Cardiol. 2003; 90. Ley S, Eichhorn J, Ley-Zaporozhan pressure gradient in children Accuracy of phase-contrast flow
91:1523–25. J, Ulmer H, Schenk JP, Kauczor HU, with coarctation of the aorta by measurements in the presence of
82. Fratz S, Hess J, Schwaiger M, Arnold R. Evaluation of aortic low-field magnetic resonance partial-volume effects. J Magn
Martinoff S, Stern HC. More regurgitation in congenital heart imaging. Pediatr Cardiol. 2002; Reson Imaging. 1993; 3:377–85.
accurate quantification of disease: value of MR imaging in 23:127–31. 108. Hofman MB, Visser FC, van
pulmonary blood flow by comparison to echocardiography. 99. Ebbers T, Wigstrom L, Bolger AF, Rossum AC, Vink QM, Sprenger M,
magnetic resonance imaging Pediatr Radiol. 2007; 37:426–36. Engvall J, Karlsson M. Estimation Westerhof N. In vivo validation of
than by lung perfusion 91. Sondergaard L, Lindvig K, of relative cardiovascular magnetic resonance blood volume
scintigraphy in patients with Hildebrandt P, Thomsen C, pressures using time-resolved flow measurements with limited
fontan circulation. Circulation. Stahlberg F, Joen T, Henriksen O. three-dimensional phase spatial resolution in small vessels.
2002; 106:1510–13. Quantification of aortic contrast MRI. Magn Reson Med. Magn Reson Med. 1995; 33:778–
83. Roman KS, Kellenberger CJ, regurgitation by magnetic 2001; 45:872–79. 84.
Farooq S, MacGowan CK, Gilday resonance velocity mapping. Am 100. Oshinski JN, Parks WJ, Markou 109. Chatzimavroudis GP, Oshinski JN,
DL, Yoo SJ. Comparative imaging Heart J. 1993; 125:1081–90. CP, Bergman HL, Larson BE, Ku Franch RH, Pettigrew RI, Walker
of differential pulmonary blood 92. Niezen RA, Helbing WA, van der DN, Mukundan S Jr, Pettigrew RI. PG, Yoganathan AP. Quantification
flow in patients with congenital Wall EE, van der Geest RJ, Improved measurement of of the aortic regurgitant volume
heart disease: magnetic Rebergen SA, de Roos A. pressure gradients in aortic with magnetic resonance phase
resonance imaging versus lung Biventricular systolic function and coarctation by magnetic velocity mapping: a clinical
investigation of the importance Roos A. MR velocity mapping of phasecontrast MRI. AJR Am J resonance perfusion. Circulation.
of imaging slice location. J Heart tricuspid flow: correction for Roentgenol. 2012; 198:W250– 2004; 110:835–42.
Valve Dis. 1998; 7:94–101. through-plane motion. J Magn 259. 133. Greenwood JP, Maredia N,
110. Kilner PJ, Gatehouse PD, Firmin Reson Imaging. 1997; 7:669–73. 127. Hsiao A, Lustig M, Alley MT, Younger JF, Brown JM, Nixon J,
DN. Flow measurement by 119. Westenberg JJ, Roes SD, Ajmone Murphy MJ, Vasanawala SS. Everett CC, Bijsterveld P, Ridgway
magnetic resonance: a unique Marsan N, Binnendijk NM, Evaluation of valvular JP, Radjenovic A, Dickinson CJ, et
asset worth optimising. J Doornbos J, Bax JJ, Reiber JH, de insufficiency and shunts with al. Cardiovascular magnetic
Cardiovasc Magn Reson. Roos A, van der Geest RJ. Mitral parallel-imaging resonance and single-photon
2007; 9:723–28. valve and tricuspid valve blood compressedsensing 4D phase- emission computed tomography
111. Kozerke S, Schwitter J, Pedersen flow: accurate quantification with contrast MR imaging with for diagnosis of coronary heart
EM, Boesiger P. Aortic and mitral 3D velocity-encoded MR imaging stereoscopic 3D velocityfusion disease (CE-MARC): a prospective
regurgitation: quantification with retrospective valve tracking. volume-rendered visualization. trial.
using moving slice velocity Radiology. 2008; 249:792–800. Radiology. 2012; 265:87–95. Lancet. 2012; 379:453–60.
mapping. 120. Gatehouse PD, Rolf MP, 128. Roes SD, Hammer S, van der 134. Jahnke C, Nagel E, Gebker R,
J Magn Reson Imaging. Graves MJ, Hofman MB, Totman J, Geest RJ, Marsan NA, Bax JJ, Kokocinski T, Kelle S, Manka R,
2001; 14:106–12. Werner B, Quest RA, Liu Y, von Lamb HJ, Reiber JH, de Roos A, Fleck E, Paetsch I. Prognostic value
112. Myerson SG. Heart valve Spiczak J, Dieringer M, et al. Flow Westenberg JJ. Flow assessment of cardiac magnetic resonance
disease: investigation by measurement by cardiovascular through four heart valves stress tests: adenosine stress
cardiovascular magnetic magnetic resonance: a multi- simultaneously using 3- perfusion and dobutamine stress
resonance. J Cardiovasc Magn centre multi-vendor study of dimensional 3-directional wall motion imaging. Circulation.
Reson. 2012; 14:7. background phase offset errors velocity-encoded magnetic 2007; 115:1769–76.
113. Cawley PJ, Maki JH, Otto CM. that can compromise the accuracy resonance imaging with 135. Ingkanisorn WP, Kwong RY,
Cardiovascular magnetic of derived regurgitant or shunt retrospective valve tracking in Bohme NS, Geller NL, Rhoads KL,
resonance imaging for valvular flow measurements. J Cardiovasc healthy volunteers and patients Dyke CK, Paterson DI, Syed MA,
heart disease: technique and Magn Reson. 2010; 12:5. with valvular regurgitation. Aletras AH, Arai AE. Prognosis of
validation. Circulation. 2009; 121. Pelc NJ. Flow quantification and Invest Radiol. negative adenosine stress
119:468–78. analysis methods. Magn Reson 2009; 44:669–75. magnetic resonance in patients
114. Gelfand EV, Hughes S, Hauser Imaging Clin N Am. 1995; 3:413– 129. Brandts A, Bertini M, van Dijk EJ, presenting to an emergency
TH, Yeon SB, Goepfert L, 24. Delgado V, Marsan NA, van der department with chest pain. J Am
Kissinger KV, Rofsky NM, 122. Chernobelsky A, Shubayev O, Geest RJ, Siebelink HM, de Roos Coll Cardiol. 2006; 47:1427–32.
Manning WJ. Severity of mitral Comeau CR, Wolff SD. Baseline A, Bax JJ, Westenberg JJ. Left 136. Buechel ER, Balmer C, Bauersfeld
and aortic regurgitation as correction of phase contrast ventricular diastolic function U, Kellenberger CJ, Schwitter J.
assessed by cardiovascular images improves quantification of assessment from three- Feasibility of perfusion
magnetic resonance: optimizing blood flow in the great vessels. J dimensional three-directional cardiovascular magnetic
correlation with Doppler Cardiovasc Magn Reson. 2007; velocityencoded MRI with resonance in paediatric patients.
echocardiography. J Cardiovasc 9:681–85. retrospective valve tracking. J J Cardiovasc Magn Reson.
Magn Reson. 2006; 8:503–07. 123. Miller TA, Landes AB, Moran AM. Magn Reson Imaging. 2009; 11:51.
115. Hundley WG, Li HF, Willard JE, Improved accuracy in flow 2011; 33:312–19. 137. Taylor AM, Dymarkowski S, De
Landau C, Lange RA, Meshack mapping of congenital heart 130. Schwitter J, Wacker CM, van Meerleer K, Hamaekers P, Gewillig
BM, Hillis LD, Peshock RM. disease using stationary phantom Rossum AC, Lombardi M, Al- M, Mertens L, Bogaert J.
Magnetic resonance imaging technique. Saadi N, Ahlstrom H, Dill T, Validation and application of
assessment of the severity of J Cardiovasc Magn Reson. Larsson HB, Flamm SD, single breath-hold cine cardiac MR
mitral regurgitation. Comparison 2009; 11:52. Marquardt M, Johansson L. MR- for ventricular function
with invasive techniques. 124. Holland BJ, Printz BF, Lai WW. IMPACT: comparison of assessment in children with
Circulation. Baseline correction of phase- perfusion-cardiac magnetic congenital heart disease at rest
1995; 92:1151–58. contrast images in congenital resonance with single-photon and during adenosine stress. J
116. Rebergen SA, Helbing WA, van cardiovascular magnetic emission computed tomography Cardiovasc Magn Reson. 2005;
der Wall EE, Maliepaard C, Chin resonance. J Cardiovasc Magn for the detection of coronary 7:743–51.
JG, de, Roos A. MR velocity Reson. 2010; 12:11. artery disease in a multicentre, 138. Manso B, Castellote A, Dos L,
mapping of tricuspid flow in 125. van der Hulst AE, Westenberg JJ, multivendor, randomized trial. Casaldaliga J. Myocardial
healthy children and in patients Kroft LJ, Bax JJ, Blom NA, de Roos Eur Heart J. 2008; 29:480–89. perfusion magnetic resonance
who have undergone Mustard or A, Roest 131. Nandalur KR, Dwamena BA, imaging for detecting coronary
Senning repair. AA. Tetralogy of fallot: 3D Choudhri AF, Nandalur MR, function anomalies in
Radiology. 1995; 194:505– velocity-encoded MR imaging Carlos RC. Diagnostic asymptomatic paediatric patients
12. for evaluation performance of stress cardiac with a previous arterial switch
117. Fujita N, Chazouilleres AF, of right ventricular valve flow magnetic resonance imaging in operation for the transposition of
Hartiala JJ, O’Sullivan M, and diastolic function in the detection of coronary artery great arteries. Cardiol Young.
Heidenreich P, Kaplan JD, patients after correction. disease: a meta-analysis. J Am 2010; 20:410–17.
Radiology. 2010; 256:724–34. Coll Cardiol. 2007; 50:1343–53. 139. Prakash A, Powell AJ,
Sakuma H, Foster E, Caputo GR,
Higgins CB. Quantification of 126. Hsiao A, Lustig M, Alley MT, 132. Paetsch I, Jahnke C, Wahl A, Krishnamurthy R, Geva T.
Murphy M, Chan FP, Herfkens RJ, Gebker R, Neuss M, Fleck E, Magnetic resonance imaging
mitral regurgitation by velocity-
Vasanawala SS. Rapid pediatric Nagel E. Comparison of evaluation of myocardial perfusion
encoded cine nuclear magnetic
cardiac assessment of flow and dobutamine stress magnetic and viability in congenital and
resonance imaging. J Am Coll
ventricular volume with resonance, adenosine stress acquired pediatric heart disease.
Cardiol. 1994; 23:951–58.
compressed sensing parallel magnetic resonance, and Am J Cardiol. 2004; 93:657–61.
118. Kayser HW, Stoel BC, van der
imaging volumetric cine adenosine stress magnetic
Wall EE, van der Geest RJ, de
140. Cook SC, Ferketich AK, Raman the American Heart Association. transposition of the great stenosis. Am J Cardiol. 2005;
SV. Myocardial ischemia in Circulation. 2002; 105:539–42. arteries. Circulation. 2005; 96:1568–70.
asymptomatic adults with 149. Arai AE. Magnetic resonance first- 111:2091–98. 164. Stranzinger E, Ensing GJ,
repaired aortic coarctation. Int J pass myocardial perfusion 157. Rathod RH, Prakash A, Powell AJ, Hernandez RJ. MR findings of
Cardiol. 2009; 133:95–101. imaging. Geva T. Myocardial fibrosis endocardial fibroelastosis in
141. Strigl S, Beroukhim R, Valente Top Magn Reson Imaging. identified by cardiac magnetic children. Pediatr Radiol. 2008;
AM, Annese D, Harrington JS, 2000; 11:383–98. resonance late gadolinium 38:292–96.
Geva T, Powell AJ. Feasibility of 150. Amado LC, Gerber BL, Gupta SN, enhancement is associated with 165. Harris MA, Johnson TR, Weinberg
dobutamine stress Rettmann DW, Szarf G, Schock R, adverse ventricular mechanics PM, Fogel MA. Delayed-
cardiovascular magnetic Nasir K, Kraitchman DL, Lima JA. and ventricular tachycardia late enhancement cardiovascular
resonance imaging in children. J Accurate and objective infarct after Fontan operation. J Am Coll magnetic resonance identifies
Magn Reson Imaging. 2009; sizing by contrastenhanced Cardiol. 2010; 55:1721–28. fibrous tissue in children after
29:313–19. magnetic resonance imaging in a 158. Giardini A, Lovato L, Donti A, surgery for congenital heart
142. Kellman P, Arai AE. Imaging canine myocardial infarction Formigari R, Oppido G, Gargiulo disease. J Thorac Cardiovasc Surg.
sequences for first pass model. J Am Coll Cardiol. 2004; G, Picchio FM, Fattori R. Relation 2007; 133:676–81.
perfusion –a review. 44:2383–89. between right ventricular 166. Fratz S, Hauser M, Bengel FM,
J Cardiovasc Magn Reson. 151. Kehr E, Sono M, Chugh SS, structural alterations and Hager A, Kaemmerer H, Schwaiger
2007; 9:525–37. Jerosch-Herold M. Gadolinium- markers of adverse clinical M, Hess J, Stern HC. Myocardial
143. Kramer CM, Barkhausen J, enhanced magnetic resonance outcome in adults with systemic scars determined by delayed-
Flamm SD, Kim RJ, Nagel E. imaging for detection and right ventricle and either enhancement magnetic resonance
Standardized cardiovascular quantification of fibrosis in human congenital complete (after imaging and positron emission
magnetic resonance imaging myocardium in vitro. Int J Senning operation) or tomography are not common in
(CMR) protocols, society for Cardiovasc Imaging. 2008; 24:61– congenitally corrected right ventricles with systemic
cardiovascular magnetic 8. transposition of the great function in long-term follow up.
resonance: board of trustees 152. Wagner A, Mahrholdt H, Holly TA, arteries. Am J Cardiol. Heart. 2006; 92:1673–77.
task force on standardized Elliott MD, Regenfus M, Parker M, 2006; 98:1277–82. 167. Kuhl HP, Papavasiliu TS, Beek AM,
protocols. J Cardiovasc Magn Klocke FJ, Bonow RO, Kim RJ, Judd 159. La Salvia EA, Gilkeson RC, Dahms Hofman MB, Heusen NS, van
Reson. 2008; 10:35. RM. Contrast-enhanced MRI and BB, Siwik E. Delayed contrast Rossum AC. Myocardial viability:
144. Gebker R, Schwitter J, Fleck E, routine single photon emission enhancement magnetic rapid assessment with delayed
Nagel E. How we perform computed tomography (SPECT) resonance imaging in congenital contrast-enhanced MR imaging
myocardial perfusion with perfusion imaging for detection of aortic stenosis. with three-dimensional inversion-
cardiovascular magnetic subendocardial myocardial Pediatr Cardiol. 2006; recovery prepared pulse
resonance. J Cardiovasc Magn infarcts: an imaging study. Lancet. 27:388–90. sequence. Radiology. 2004;
Reson. 2007; 9:539–47. 2003; 361:374–79. 160. Robinson JD, Del Nido PJ, Geggel 230:576–82.
145. Gerber BL, Raman SV, Nayak K, 153. Babu-Narayan SV, Kilner PJ, Li W, RL, Perez-Atayde AR, Lock JE, 168. Foo TK, Stanley DW, Castillo E,
Epstein FH, Ferreira P, Axel L, Moon JC, Goktekin O, Davlouros Powell AJ. Left ventricular Rochitte CE, Wang Y, Lima JA,
Kraitchman DL. Myocardial first- PA, Khan M, Ho SY, Pennell DJ, diastolic heart failure in Bluemke DA, Wu KC. Myocardial
pass perfusion cardiovascular Gatzoulis MA. Ventricular fibrosis teenagers who underwent viability: breath-hold 3D MR
magnetic resonance: history, suggested by cardiovascular balloon aortic valvuloplasty in imaging of delayed
theory, and current state of the magnetic resonance in adults with early infancy. Am J Cardiol. 2010; hyperenhancement with variable
art. J Cardiovasc Magn Reson. repaired tetralogy of fallot and its 106:426–29. sampling in time. Radiology. 2004;
2008; 10:18. relationship to adverse markers of 161. Liang XC, Lam WW, Cheung EW, 230:845–51.
146. Jerosch-Herold M. Quantification clinical outcome. Wu AK, Wong SJ, Cheung YF. 169. Kellman P, Arai AE, McVeigh ER,
of myocardial perfusion by Circulation. 2006; 113:405– Restrictive right ventricular Aletras AH. Phase-sensitive
cardiovascular magnetic 13. physiology and right ventricular inversion recovery for detecting
resonance. J Cardiovasc Magn 154. Oosterhof T, Mulder BJ, Vliegen fibrosis as assessed by cardiac myocardial infarction using
Reson. 2010; 12:57. 147. HW, De RA. Corrected tetralogy of magnetic resonance and gadolinium-delayed
Klem I, Heitner JF, Fallot: delayed enhancement in exercise capacity after hyperenhancement. Magn Reson
Shah DJ, Sketch MH Jr, Behar V, right ventricular outflow tract. biventricular repair of Med. 2002; 47:372–83.
Weinsaft J, Cawley P, Parker M, Radiology. 2005; 237:868–71. pulmonary atresia and intact 170. Desai MY, Gupta S, Bomma C,
Elliott M, Judd RM, Kim RJ. 155. Wald RM, Haber I, Wald R, Valente ventricular septum. Clin Cardiol. Tandri H, Foo TK, Lima JA,
Improved detection of coronary AM, Powell AJ, Geva T. Effects of 2010; 33:104–10. Bluemke DA. The apparent
artery disease by stress regional dysfunction and late 162. Fratz S, Hager A, Schreiber C, inversion time for optimal delayed
perfusion cardiovascular gadolinium enhancement on Schwaiger M, Hess J, Stern HC. enhancement magnetic resonance
magnetic resonance with the use global right ventricular function Long-term myocardial scarring imaging differs between the right
of delayed enhancement and exercise capacity in patients after operation for anomalous and left ventricles.
infarction imaging. J Am Coll with repaired tetralogy of Fallot. left coronary artery from the J Cardiovasc Magn Reson.
Cardiol. 2006; 47:1630–38. Circulation. 2009; 119:1370–77. pulmonary artery. Ann Thorac 2005; 7:475–79.
148. Cerqueira MD. Standardized 156. Babu-Narayan SV, Goktekin O, Surg. 2011; 92:1761–65. 171. Amano Y, Kumazaki T. Differences
Myocardial Segmentation and Moon JC, Broberg CS, Pantely GA, 163. Tworetzky W, del Nido PJ, Powell in null points between the left and
Nomenclature for Tomographic Pennell DJ, Gatzoulis MA, Kilner AJ, Marshall AC, Lock JE, Geva T. right ventricles in contrast-
Imaging of the Heart: A PJ. Late gadolinium enhancement Usefulness of magnetic enhanced inversion recovery MR
Statement for Healthcare cardiovascular magnetic resonance imaging of left imaging in patients with cardiac
Professionals From the Cardiac resonance of the systemic right ventricular endocardial diseases. Comput Med Imaging
Imaging Committee of the ventricle in adults with previous fibroelastosis in infants after Graph. 2006; 30:147–51.
Council on Clinical Cardiology of atrial redirection surgery for fetal intervention for aortic valve
172. Grosse-Wortmann L, Macgowan

CK, Vidarsson L, Yoo SJ. Late
gadolinium enhancement of the
right ventricular myocardium: is
it really different from the left? J
Cardiovasc Magn Reson. 2008;
10:20.
173. Hsu LY, Natanzon A, Kellman P,
Hirsch GA, Aletras AH, Arai AE.
Quantitative myocardial
infarction on delayed
enhancement MRI. Part I:
Animal validation of an
automated feature analysis and
combined thresholding infarct
sizing algorithm. J Magn
ResonImaging. 2006; 23:298–
308.
174. Kim RJ, Shah DJ, Judd RM. How
imaging. J Cardiovasc Magn Reson. 2003 ; 5:505 –14.
doi:10.1186/1532-429X-15-51
Cite this article as:Fratz et al.: Guidelines and protocols for
cardiovascular magnetic resonance in children and adults with
congenital heart disease: SCMR expert consensus group on congenital
heart disease.Journal of Cardiovascular Magnetic Resonance2013 15:51.
Submit your next manuscript to BioMed Central

and take full advantage of:
• Convenient online submission

• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at

www.biomedcentral.com/submit
we perform delayed
enhancement

Anak 3

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Anak 3

Uploaded by

Copyright:

Available Formats

Fratz et al.

Journal of Cardiovascular Magnetic Resonance 2013, 15:51 http://jcmr-

POSITION STATEMENT Open Access

Guidelines and protocols for cardiovascular

Introduction preparation for the examination and safety issues, the

For rates greater than 100 bpm, it may be helpful to

Respiratory compensation Free-breathing

Number of signal averages 3

Reconstructed phases per R-R interval

Isotropic resolution (mm3) 1.2-1.5 1.3-2.0

Image acquisition duration (ms) 40-60 80-150

Figure 5 Coronary MR angiography. Patient with anomalous

In addition to the technical points mentioned in the

in both the short-axis and axial planes, or two sets of

measurement. A position image. Again, through-

drawn on the outer margin motion of the heart and

animal models and in perform a MRA. For gadolinium concentration,

right ventricular encompass both

3. CE-MRA or 3D parameters including e, and systemic Standard

1. Cine CMR: axial 2. Ventriculography module with multiplecontiguous

the right pulmonary native MPA,

172. Grosse-Wortmann L, Macgowan

Submit your next manuscript to BioMed Central

• Convenient online submission

Submit your manuscript at

You might also like