| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Original Research |
1 Department of Radiology, Cairo University, Kasr Al-Ainy Hospital, 4 49th St., Cairo 11571, Egypt.
Received February 14, 2008;
accepted after revision April 24, 2008.
Address correspondence to S. N. Saleem
(saharsaleem1{at}gmail.com).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. After technically inadequate echocardiography, MRI was performed on 20 fetuses (mean gestational age, 24 weeks; range, 18–32 weeks) at risk of congenital heart disease. MRI was attempted along the three fetal body planes (n = 20) and cardiac axes (n = 3) without fetal sedation. The images were analyzed with an anatomic segmental approach. Each feature was classified as well visualized or poorly or not visualized. In each group, the Student's t test was used to assess the relation between visibility of fetal cardiac features and gestational age.
RESULTS. Imaging was possible along the fetal body and cardiac axes. In the axial plane, a balanced four-chamber view was obtained in all fetuses, enabling evaluation of heart position, axis, chambers, and interventricular septum. The left and right ventricular outflow tracts were well visualized in 12 (60%) and nine (45%) of the fetuses, respectively; the three-vessel view was obtained in 10 fetuses (50%). With the combination of sagittal and coronal views, both ventricular outflow tracts were assessed in all fetuses. The superior and inferior venae cavae were identified in all fetuses, and at least one pulmonary vein was visualized in 17 fetuses (85%). There were no statistically significant differences between gestational age and lack of visualization of a cardiac feature that was attributed to fetal motion.
CONCLUSION. MRI of the fetal heart with a steady-state free precession sequence in multiple planes and image analysis with an anatomic segmental approach to congenital heart disease are possible in situations that limit echocardiography.
Keywords: fetus • heart • MRI
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
MRI does not have the limitations of sonography [6]. Studies [6, 7] have shown that fetal MRI is particularly helpful in evaluation of the CNS and other fetal body parts. Ex vivo fetal MRI of the cardiac structures has better image quality and structural detail in comparison with echocardiography [8]. However, MRI of the cardiovascular system has not been thoroughly investigated in utero because of the technical difficulties caused by fetal cardiac motion, fetal body motion, inevitable maternal motion such as peristalsis and respiration, and effects of blood flow [7]. A small number of case reports describe MRI of cardiac structures in utero. Most of these reports are limited to images obtained with a single-shot fast spin-echo (SSFSE) sequence, which show the heart and vessels as signal-void structures with limited spatial resolution [9–11]. As a consequence, there remains room for use of other sequences in studying the fetal cardiovascular system. Balanced MRI sequences such as steady-state free precession (SSFP) improve blood-pool homogeneity and allow faster acquisition, which results in better image resolution and fewer motion artifacts [12–14]. Two case reports [15] and a retrospective study [16] of fetal body MRI of 10 fetuses have provided preliminary evidence of the feasibility of imaging the heart in utero with an SSFP sequence.
To my knowledge, the literature does not contain a report of a prospective study of MRI of the heart in utero along body and cardiac axes. The aim of this work was to evaluate the feasibility of imaging the fetal heart with an SSFP sequence along body and cardiac axes after inadequate echocardiography and the feasibility of analyzing these images with an anatomic segmental approach to congenital heart disease.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Imaging
The study had institutional review board approval, and written informed
consent was obtained from all of the expectant mothers before MRI. The
expectant parents and their referring physicians were informed that fetal MRI
was investigative and of unproven accuracy in the diagnosis of cardiac
malformations. No patients were excluded because of contraindications to MRI.
MRI was performed with a 1.5-T superconducting magnet (Gyroscan Intera,
Philips Healthcare) and a phased-array surface coil. Balanced SSFP sequences
(TR/TE, 3.5–4/1.7–2; flip angle, 60–90°; slice
thickness, 4–6 mm; gap, 1 mm to section overlap of –3 mm) were
performed for all patients. The thin overcontiguous sections were used for
evaluation of small fetuses. The field of view (205–350 mm2)
was adjusted to increases or decreases in the fetal or maternal dimensions or
when aliasing artifact was a problem. The matrix size was 183 x 256 or
256 x 256 according to the selected field of view; reduction of field of
view compelled reduction of matrix size in some cases. One to three signals
were averaged. The imaging time for each sequence depended on the number of
slices and number of signals averaged; typically, 16 seconds was needed to
obtain 16 slices with two signals averaged.
Certain clinical measures were taken to reduce motion artifacts during the study. The pregnant woman was advised not to have anything by mouth for at least 4 hours before image acquisition to prevent postprandial motion. The mother was advised to empty her bladder before the study to feel more comfortable. All of the expectant mothers tolerated the examination in the supine position. After a scout acquisition, a series of images along the three fetal body planes (axial, sagittal, and coronal) were obtained for each patient. Each new acquisition was prescribed by use of images from the immediately previous acquisition to avoid misregistration caused by fetal movement. In most cases, the thoracic and extrathoracic fetal anatomic features were adequately identified on fetal planar images. Extracardiac fetal parts were studied only in selected cases in which sonographic findings were inadequate. These additional images were obtain ed along the axial planes of the brain and abdo men in six cases. Controlled maternal breath-holding was useful in decreasing motion artifacts in some cases; however, SSFP was acquired during free maternal breathing in most cases.
In three arbitrarily selected cases, acquisition of additional MR images was attempted along cardiac axes. A representative image from each sequence was used as a scout to align the subsequent acquisition. From the coronal plane, the long-axis view was obtained along a line that extended from the cardiac apex to the middle of the left ventricle. A short-axis view perpendicular to the long axis also was obtained. In the short-axis view, the image in which the left ventricle was concentric and the right ventricle was crescent-shaped was chosen to align the plane for the four-chamber view. The plane for the four-chamber view extended along the center of the left ventricle and the farthest corner of the right ventricle. From the four-chamber image, the two-chamber view was obtained along the right and left sides of the heart [17, 18].
No maternal sedation, fetal sedation or paral ysis, fetal cardiac gating, or contrast media were used in any of the cases in the study. Claus tropho bia was not a problem because the patients were adequately instructed and comforted during the examination, which lasted 15–25 min utes (aver age, 19 minutes). Evaluation of extra cardiac fetal structures in the selected cases did not markedly prolong the average study time because each additional sequence took only approximately 16 seconds.
Image Interpretation
All MRI examinations were attended and the images interpreted by the
author, who had 18 years of experience in body MRI and 10 years of experience
in fetal imaging. Data obtained at MRI included referral indication and
gestational age according to last menstrual period and earliest sonographic
biometric measurements. In the axial plane, five transverse views were sought:
upper abdomen showing the stomach, four chamber, left ventricular outflow
tract, right ventricular outflow tract, and three vessel. Each view was
classified as well visualized or poorly or not visualized. For evaluation of
MR images, fetal cardiac structures were analyzed with a modified anatomic
segmental approach to congenital heart disease
[19] that included the
following 10 points:
Visceroatrial situs—The stomach normally is on the left and the inferior vena cava on the right side of the body. Assessment of the situs in relation to the bronchi was attempted in each case. The embryologic left bronchus is long with no early division and runs under the left pulmonary artery; the right bronchus is short, closer to vertical, and runs behind the right pulmonary artery.
Cardiac position—Most of the heart normally occupies the left side of the thorax with the cardiac apex to the left.
Cardiac size—The heart normally occupies one third of the thorax.
Cardiac axis—The cardiac axis is the angle between the true sagittal plane (between the spine and the anterior chest wall) and a plane along the interventricular septum.
Cardiac chambers—The cardiac chambers were evaluated for number (normally four), arrangement (the left atrium is the chamber closest to the fetal spine in a normal heart), relative size of both atria (normally equal in size), and relative size of both ventricles (normally equal in size). Morphologic identification of the embryologic right ventricle was achieved by visualization of a trabeculated septal wall or, more specifically, the moderator band (a muscular structure that connects the apical ventricular septum with the lateral wall of the right ventricle). The following cardiac structures also were analyzed: foramen ovale, atrial septal duct valve location (presence in the left atrium indicates normal right-to-left blood flow), intactness of interventricular septum, and atrioventricular and great arterial valves.
Ventricular looping—In a normal D-loop, the more anterior ventricle is the embryologic right one.
Inflow veins—Normally, the superior and inferior venae cavae connect to the right atrium. Two right and two left pulmonary veins normally join the left atrium.
Outflow vessels—The left and right ventricular outflow tracts (aorta and pulmonary arteries) were evaluated for relative size (normally equal) and relative position in relation to each other (normally cross perpendicular in relation to their origin).
Ventriculoarterial concordance—Normally the aorta arises from the left ventricle. It can be traced in a regular arch that gives rise to three neck vessels. The pulmonary artery normally arises from a morphologic right ventricle and bifurcates at its distal end.
Side of the aortic arch—The side of the aortic arch was defined according to the main bron chus, above which it crossed the mediastinum posteriorly.
Statistical Analysis
At analysis of MR images, all cardiac anatomic components were classified
into one of two categories: well visualized or poorly or nonvisualized. In
each group, the Student's t test was used to assess the relation
between visibility of fetal cardiac features and gestational age. A value of
p < 0.05 was considered to indicate a statistically significant
difference.
Outcome Measures
The patients in the study underwent prenatal evaluation of their clinical
course and postnatal clinical and cardiac follow-up.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
|
|
|
The left (Fig. 1B) and right (Fig. 1C) ventricular outflow tracts were identified in their corresponding transverse views in 12 (60%) and nine (45%) fetuses, respectively. The coronal (Fig. 2A, 2B) and sagittal (Figs. 3A, 3B, 3C and 4) planes showed the left and right ventricular outflow tracts, respectively, in all of the imaged fetuses. With the combination of sagittal and coronal views, both ventricular outflow tracts were assessed for relative size and crossing at their origin in all cases. A normal visceroatrial situs in relation to the bronchi was at the right ventricular outflow tract level (Fig. 1C).
|
The inflow veins—superior and inferior venae cavae—were identified in all cases in the sagittal (Fig. 3A) and coronal (Fig. 2A) planes. In 17 fetuses (85%), at least one pulmonary vein (the left) was identified in the axial plane. A right pulmonary vein was visualized in 14 fetuses (70%). The valves were infrequently visualized. The atrioventricular valves were occasionally seen in the axial plane (n = 4) and the pulmonary cusps (Fig. 3C) in the sagittal plane (n = 3).
Table 1 shows the statistical relationships between gestational age and visibility of fetal cardiac features on MR images obtained along the three body planes with SSFP in 20 fetuses. There were no significant statistical difference between gestational age and lack of visualization of any of the cardiac features studied.
|
Imaging along the cardiac axes was successful in all three cases in which it was attempted. The mean gestational age was 27 weeks (range, 26–28 weeks). The morphologic details of the cardiac chambers, the moderator band (Fig. 5A), and the atrial septal flap (Fig. 5D) were well identified in the long-axis and four-chamber view in all three cases. The pulmonary veins were well identified in the long-axis and four-chamber views (Fig. 5A). The ventricular outflow tracts were well displayed on images obtained along the short cardiac axis (Figs. 5B and 5C). The atrioventricular connections were well seen in the two-chamber view (Fig. 5E) in all three cases studied.
|
|
|
|
|
No congenital heart defects or extracardiac anomalies were diagnosed with fetal MRI in the cases studied. The expectant parents and their referring physicians were informed of the results of fetal cardiac MRI with the caveat that the examination was investigative and of unproven accuracy for the diagnosis of heart malformations. At the time of this writing, all 20 fetuses continued in normal pregnancies and were alive with normal results of physical and postnatal cardiac sonographic examinations.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SSFP sequences entail use of balanced-gradient waveforms, which act on stationary spins on resonance between two consecutive radiofrequency pulses and return them to the same phase they had before the gradients were applied [12]. Because balanced gradients maintain the longitudinal and transverse magnetizations, both T1- and T2-weighted tissue contrast are represented in the resultant image. The SSFP sequence produces images with increased signal intensity from fluid (like T2-weighted images) and retains T1-weighted tissue contrast. The result is enhanced tissue contrast [13]. Moreover, SSFP is partially flow-compensated, thus the blood pool in the cardiac chambers and vessels has homogeneous high signal intensity [17]. Visualization of the cardiac chambers and vessels on SSFP images is better than on SSFSE images, on which the chambers have low or variable signal intensity [16]. SSFP images have high T2-weighted contrast. Although a relatively larger field of view is needed, the SSFP sequence allows use of a very thin slice thickness with a high signal-to-noise ratio over contiguous slices for studies of the fetal thorax and great vessels [14, 16].
This study was conducted with the SSFP sequence for study of fetal cardiac structures after inadequate echocardiography. The results show that MRI of the heart in utero is possible with SSFP after technically limited echocardiography. Although artifacts due to fetal movement continue to be a concern in MRI, the very short acquisition time with the SSFP sequence in association with the mother's cooperation resulted in overall good image quality in this study without mater nal sedation or fetal paralysis or sedation. MRI of the fetus also did not include special techniques such as fetal cardiac gating and maternal breath-holding. Diagnostic views orthogonal to the three body axes and along the cardiac axes were obtained. Successful analysis of the fetal cardiac images was possible through a modified segmental approach to congenital heart disease [19].
Cardiac MRI along the three body axes (transverse, coronal, and sagittal) is appropriate for evaluation of the overall structure of the heart and great vessels [18]. Because situs abnormalities often are associated with major cardiac malformations, fetal heart screening usually begins with identification of the fetal situs and the arrangement of the heart in the transverse plane [2, 3]. In this study, MRI showed the visceroatrial situs in all of the fetuses by showing the stomach in multiple planes. The stomach and the heart were on the same side of the body in all of the fetuses. Unlike sonography [16], fetal MRI allowed assessment of the visceroatrial situs in relation to the bronchi. The fluid-filled bronchial tree appears as high-signal-intensity structures on SSFP images [14]. The morphologic features of the embryologic right and left bronchi were clearly seen in relation to the pulmonary artery divisions in multiple planes in all of the fetuses in this study.
Cross-sectional imaging of the heart in the axial plane is useful for studying the cardiac chambers and the pericardium [3, 7, 18]. In many instances, only a four-chamber transverse view of the heart is obtained in sonographic fetal screening [4]. In a retrospective analysis [16] of body MRI of 10 fetuses, a balanced four-chamber view of the heart was feasible in all cases. In my prospective study, a four-chamber transverse view of the heart was obtained in all cases. The four-chamber transverse view provided valuable information about the size, position, and axis of all of the hearts studied. The cardiac axis measurement in this study (37.25° ± 7.15°) is comparable with sonographic norms (43° ± 2°)[20]. When the axis of the heart is not within the normal range, the fetus is at increased risk of heart malformation or abnormal intrathoracic anatomy [2]. The four-chamber view depicted four balanced cardiac chambers in all of the cases studied. Both atria and both ventricles were equal in size, and the interventricular septum was central and intact. These findings allowed exclusion of an odd number of chambers; unbalanced abnormalities of atrioventricular valves, such as atresia; and the spectrum of atrioventricular septal defects.
Although visualization of the detailed structure of the interatrial septum—the foramen ovale and the atrial septum duct valve (flap) within the left atrium—was possible in the transverse plane in some cases, these structures were better evaluated along cardiac axes (long-axis and four-chamber views). Identification of the flap in the left atrium indicated normal right-to-left blood shunt across the foramen ovale during fetal life and excluded heart anomalies that cause forward flow to reverse, such as transposition of the heart chambers and obstruction of the flow of the left side of the heart [21].
Recognition of a moderator band or trabeculated septal wall helped in identification of the embryologic right ventricle in 65% of cases. Recognition of ventricular morphologic features allows determination of ventricular looping and analysis of atrioventricular and ventriculoarterial concordance [18].
The screening sensitivity of the four-chamber view in echocardiography was estimated to be only 60–64% for cardiac malformations [4, 22]. However, visualization of the right and left ventricular outflow tracts in other views was estimated to increase the sensitivity of sonographic examination to 75–92% in the detection of congenital cardiac defects in utero [4, 22, 23]. In my study, the left and right outflow tracts were visualized in 60% and 45% of cases, respectively, when the axial MR images alone were analyzed. In the combination of sagittal and coronal images with axial images, however, both ventricular outflow tracts were visualized in all fetuses. Images in the coronal plane depicted the left ventricular outflow tract, and images in the sagittal plane, the right ventricular outflow tract. The relative size of the fetal great vessels and their characteristic branching pattern were clearly assessed in multiple MRI planes. Identification of the normal crossing of the aorta and pulmonary arteries at their origins was important for excluding transposition of the great arteries. Because of its multiplanar capability along body planes and cardiac axes, MRI well depicted the systemic and pulmonary venous connections of the fetal heart. The findings suggest a potential role in the diagnosis of pathologic changes in the cardiac venous inlet.
There was no significant statistical correlation between gestational age and visualization of a cardiac features in this study. The poor quality of some images with less structural detail can be explained by artifacts caused by fetal motion rather than limitation of the spatial resolution of the MRI sequence used. Lack of visualization of fetal cardiac structures on MRI in a study by Gorincour et al. [16] also was attributed to fetal movement. The results of this study showed the inconsistency of visualization of important fetal cardiac anatomic features when only fetal axial planes are used and emphasize the need for conventional cardiac planar techniques. Imaging along cardiac planes has proven [2, 15, 18] ideal for identification of normal cardiac anatomic and abnormal morphologic features, particularly ventricular defects and conal–truncal abnormalities. Obtaining MR images along the short and long cardiac axes and in the four-chamber and two-chamber views was possible in this study and enabled reliable visualization of fetal cardiac structures.
Although imaging of the heart orthogonal to the chest planes can be appropriate for overall evaluation of the morphologic features of the heart, it is not suitable for quantitative measurements or acquisition of functional data [17]. Because orthogonal planes are not perpendicular to the cardiac wall or the cavity, partial volume effects and obliqueness can cause considerable overestimation of the true measurements of wall thickness and cavity dimensions. Imaging along the cardiac planes is more suitable for this purpose [18].
The feasibility of quantitative analysis of cardiac MRI in utero was evaluated by Fogel and colleagues [15], who performed real-time MRI with the SSFP sequence on two fetuses with cardiac malformations. Those investigators found that fetal MRI produced measurements of ventricular volume that could not be obtained with fetal echocardiography. Both fetuses, however, were at advanced ages of gestation (34.5 and 36 weeks), and fetal sedation was necessary in one case.
Although cine MRI with the SSFP sequence is an established method of evaluation of cardiac function [24], imaging of a beating heart in a moving fetus is more complicated. The normal fetal heart rate of 120–160 beats per minute and unpredictable fetal body motion impair image quality owing to motion artifacts during MRI acquisition in utero [24, 25]. With the advent of parallel MRI technique, acquisition time is markedly reduced, yet image quality meets clinical requirements. Studies [14, 25] have shown that cine MRI in utero with steady-state acquisition and parallel imaging may be useful in the assessment of fetal body and cardiac movements.
Future fetal MRI sequences can take advantage of the improvement in radiofrequency and computing technology to decrease acquisition time and increase signal-to-noise ratio and image resolution. Fetal cardiac gating devices also may soon be available. As MRI technology advances, images of the heart are expected to be much clearer and to be acquired in a shorter time, allowing dramatic improvement in anatomic and functional MRI of the fetal heart. Reliable cardiac imaging in utero may have implications in the management of congenital heart disease, opening the possibility of fetal surgery.
This study had limitations, such as the relatively small sample size of 20 cases and the use of imaging along cardiac planes in only three cases. Fetal cardiac structures are visualized more reliably with cardiac planar techniques, which have been used with fetal sonography and are now feasible with fetal MRI, as shown in this study.
Imaging of the fetal heart in multiple body and cardiac planes with SSFP and analyzing images with an anatomic segmental approach to congenital heart disease are feasible. Further studies are needed to evaluate the diagnostic utility and accuracy of in utero MRI of normal and abnormal hearts in correlation with the reference standard technique, fetal echocardiography.
In conculsion, fetal MRI with the SSFP sequence after inadequate echocardiography facilitates visualization of the cardiac structures in multiple planes and image analysis with the anatomic segmental approach to congenital heart diseases. These features may be helpful for further characterization of cardiovascular abnormalities in utero, especially in situations that limit echocardiography.
Acknowledgments
I thank Amal Mahmoud, the MRI operator at Cairo University.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
