Diagnostic Imaging: Echocardiography and Magnetic Resonance Imaging

CHAPTER 107 Diagnostic Imaging

Echocardiography and Magnetic Resonance Imaging

Before the advent of cardiopulmonary bypass in the mid 1950s, little attention was given to the diagnosis of congenital heart disease (CHD), because no effective treatment was available. Physical examination, auscultation, electrocardiography, and radiography were the main diagnostic tools. Progress in open-heart techniques for repair of CHD required accurate and comprehensive delineation of cardiovascular anatomy and function. During the 1960s and 1970s, cardiac catheterization and angiography were the principal tools used for diagnosis of CHD. Echocardiography entered the arena in the late 1970s. The diagnostic capability of M-mode echocardiography proved insufficient for patients with CHD, but the rapid evolution of two-dimensional (2D) echocardiography during the following decade transformed the field. The technological advances in transducer design, and in image processing and display, together with development and refinement of new imaging planes and examination techniques, allowed high-quality tomographic visualization of most cardiac defects.1 The application of Doppler ultrasound to investigate blood flow allowed comprehensive hemodynamic assessment. By the mid 1980s, much of the necessary anatomic and hemodynamic information required for patient management could be obtained noninvasively, obviating the need for a diagnostic catheterization in many patients. During the late 1990s and into the 2000s, the field of pediatric cardiac imaging experienced accelerated progress in areas such as three-dimensional (3D) echocardiography, sophisticated techniques for assessment of myocardial function, and the application of magnetic resonance imaging (MRI) to CHD. At the same time, the proportion of cardiac catheterization procedures performed solely for diagnostic purposes has drastically declined. More recently, the role of cardiac imaging for pre-intervention planning and for guidance of cardiovascular procedures has expanded dramatically.

This chapter discusses the clinical application of the two main noninvasive diagnostic imaging modalities used for anatomic and physiologic evaluation of preoperative and postoperative CHD—echocardiography and cardiovascular MRI.


Echocardiography is an ideal diagnostic tool in pediatric cardiology because of its noninvasive nature, relatively low cost, superb spatial and temporal resolutions, and ability to image cardiovascular anatomy, thus allowing evaluation of physiology in real time. In addition, modern cardiac ultrasound equipment is portable and adaptable to different environments, such as the operating room, the intensive care unit, the bedside, and the outpatient office. In today’s pediatric cardiology practice, echocardiography is the primary diagnostic modality used to evaluate anatomy and physiology preoperatively, intraoperatively, postoperatively, during follow-up of CHD, and prenatally.13


To obtain an echocardiographic image, a burst of ultrasound energy is generated by a piezoelectric crystal and travels through the soft tissue at an average speed of approximately 1540 m/sec. When the propagating ultrasound wave encounters an interface between tissues with different acoustic properties, some of the energy is reflected back toward the transducer and some of the energy is refracted and continues to travel in the medium until it encounters the next interface. The returning ultrasound energy is then converted into an electrical energy that goes through a series of electronic processes, including amplification, filtering, post-processing, and display.

Three-Dimensional Echocardiography

Accurate spatial perception of an object depends on recognition of its three dimensions: length, width, and depth. Although an experienced examiner can mentally construct a 3D image of the heart from serial 2D tomographic images obtained by sweeping the transducer across the chest, 3D echocardiography offers an enhanced perspective of cardiovascular structures and their interrelations. Previously, the approach for obtaining 3D echocardiographic images of the heart was based on computer reconstruction of contiguous 2D cross-sectional images. These efforts were hampered by difficulties in accurately registering the ultrasound image data in time and space and by long processing times. Nowadays, computer reconstruction of 2D images to produce 3D images has largely been replaced by real-time 3D echocardiography. This technology, which is based on a new generation of matrix-array transducers with several thousands of simultaneously transmitting and receiving piezoelectric elements and sophisticated parallel data processing, provides real-time 3D images with sufficient temporal resolution to display in cine-loop format (Fig. 107-3).57 More recently, miniaturization of matrix-array transducer technology has allowed development of a real-time 3D transesophageal echocardiographic probe.8 Further refinements of this technology will result in continued improvement of spatial and temporal resolutions as well as better, more intuitive, user interface. Such advances are likely to contribute to the ongoing acceptance of this technology in routine practice.

Contrast Echocardiography

As early as the late 1960s, Gramiak and colleagues11 noted that intravascular injection of almost any solution resulted in a contrast effect detectable by echocardiography. Initially, this technique was used to identify structures seen by M-mode echocardiography. Contrast echocardiography has been used to detect systemic12 and pulmonary venous anomalies,13 and for the detection of intracardiac and great artery level shunts.14 In today’s pediatric echocardiography, contrast studies are infrequently performed and are usually limited to detection of intracardiac shunts in patients with limited echocardiographic windows, patch or baffle leak after cardiac surgery, and pulmonary arteriovenous malformations.2,3

Objectives of the Echocardiographic Examination

The objectives of the echocardiographic examination must be tailored to the individual patient. The initial evaluation should include a comprehensive survey of all anatomic elements of the central cardiovascular system.15 Subsequent examinations are often targeted to answer specific clinical questions. It is important, however, to repeat complete echocardiograms during follow-up, even in patients who underwent a comprehensive initial examination, because of the dynamic nature of CHD. Examples include the late onset of discrete subaortic stenosis in patients with ventricular septal defect or coarctation of the aorta,16 double-chambered right ventricle,17 and supramitral stenosing ring.18

Examination Technique

Proper planning of the echocardiographic examination ensures that all diagnostic information is obtained most efficiently. This is particularly relevant in sedated patients, for whom the time available for data acquisition is limited. Ideally, a complete segmental examination of cardiovascular anatomy and function is performed in every new patient. This includes determination of visceral situs, heart position, atrial situs, systemic and pulmonary venous connections, ventricular situs, atrioventricular and ventriculoarterial alignments and connections, and coronary and great arterial anatomy. Assessment of ventricular function, intracardiac and vessel dimensions, and flow analyses across all valves, septa, chambers, and vessels are integral parts of the examination. In young children with suspected heart disease, the examination begins from the subxiphoid approach by determining the abdominal situs and then proceeding by scanning the heart and great vessels, employing a step-by-step segmental analysis (Fig. 107-5A,B).15,19,20 This approach is advantageous because it provides a wide-angle view of heart position and cardiovascular anatomy and function at an early stage in the examination. Subsequent 2D and Doppler analyses from the apical, parasternal, and suprasternal notch views supplement and confirm findings from the subxiphoid window (see Fig. 107-5C-G). The examination strategy should be tailored to the individual patient and modified according to the clinical situation as necessary. Although the standard views just described should be obtained in almost every patient, and represent the minimum acceptable examination, flexibility and improvisation are important to optimally use the full potential of echocardiography.

image image

Figure 107–5 Standard two-dimensional transthoracic imaging sweeps. A, Subxiphoid long-axis sweep. Slow gradual sweep starting at the level of the upper abdomen will show the connection of the inferior vena cava to the right atrium (RA). The left atrium (LA) is seen next. The connection of the pulmonary veins and the atrial septum can be demonstrated from this view. The left ventricle (LV) is seen along its long axis. Further superior angulation of the transducer depicts the left ventricular outflow tract, aortic valve, and ascending aorta (Ao). The superior vena cava (SVC) is seen to the right of the ascending aorta and the main pulmonary artery (MPA) is seen to the left of the aorta. Further superior tilt of the transducer shows the right ventricular inflow (RV inflow) and outflow (RV outflow) and the pulmonary valve. The sweep ends with the anterior free wall of the right ventricle. B, Subxiphoid short-axis sweep. From the subxiphoid long-axis view, the transducer is rotated clockwise about 90 degrees. The sweep begins at the rightward-most aspect of the heart and progresses from right to left through the cardiac apex. The superior vena cava (SVC) and inferior vena cava are seen entering the right atrium (RA). The right pulmonary artery (RPA) is seen in cross section behind the SVC and above the left atrium (LA). The atrial septum is well seen in this plane. Sweeping the transducer leftward shows the base of the left ventricle (LV) and right ventricle (RV) and the AV valves. The aortic (Ao) valve is seen in cross section at this level. Further leftward tilt of the transducer depicts a cross-sectional view of the LV and mitral valve (MV), as well as the right ventricular outflow tract and pulmonary valve (PV). The sweep ends with imaging of the mid-muscular septum, the papillary muscles, and the apical portions of both ventricles. C, Apical four-chamber sweep. The transducer is positioned over the apex and angled to obtain a cross-sectional view of the atria and ventricles as shown in level 2. The transducer is then angled posteriorly to image the posterior aspect of the heart (level 3). In this plane, the coronary sinus (CS) can be viewed along the posterior left atrioventricular groove. Anterosuperior tilt of the transducer will show the left ventricular outflow tract and proximal ascending aorta (Ao). D, Parasternal long-axis sweep. The transducer is placed over the left precordium to the left of the sternum with the index mark toward the patient’s right shoulder. A rightward and inferior tilt of the transducer toward the right hip shows the right atrium (RA), tricuspid valve, and right ventricular inflow (RV) (level 1). The coronary sinus can be followed into the right atrium in this view. A leftward and superior tilt of the transducer toward the left shoulder depicts the right ventricular outflow tract (RV), pulmonary valve, and main pulmonary artery (PA) (level 3).

E, Parasternal short-axis sweep. From the parasternal long-axis view, the transducer is rotated clockwise about 90 degrees. The sweep progresses from a plane that shows the right and left atria (LA and RA), atrial septum, tricuspid valve (TV), right ventricle (RV), pulmonary valve (PV), and main pulmonary artery (level 1), toward the apex. Cross-sectional views of the right ventricle (RV), left ventricle (LV), ventricular septum, mitral valve (MV), and papillary muscles (PMs) are obtained. (F) Aortic arch view from the suprasternal notch window. The innominate vein (Innom. Vein) is seen anterior to the innominate artery. The right pulmonary artery (RPA) is seen in cross section behind the ascending aorta (Ao). G, Suprasternal notch view in the transverse plane. The left innominate vein (Innom. V) is seen draining into the superior vena cava (SVC). The distal ascending aorta (Ao) is seen superior to the right pulmonary artery (RPA), which is seen along its length above the left atrium (LA). Note the pulmonary veins entering the left atrium. H, High right parasternal view in the sagittal plane view shows the superior vena cava (SVC) entering the right atrium (RA). This view allows demonstration of the sinus venosus septum. RUPV, right upper pulmonary vein.

Special Echocardiographic Procedures

Transesophageal Echocardiography

Transesophageal echocardiography (TEE) was introduced in 1976 and appeared in pediatric use in 1989. The miniaturization of probes and development of multiplanar imaging have greatly increased its role as an adjunct to transthoracic imaging, during surgical repair of CHD (intraoperative TEE), and to guide interventional catheterization procedures. Advances in transducer technology and image processing now allow real-time 3D TEE.8

Indications and Objectives

A TEE examination is usually performed to answer a limited set of clinical questions. It is advisable, however, to perform a comprehensive examination of the heart and blood vessels for additional unsuspected anatomic or hemodynamic anomalies (Fig. 107-6).21 Miniaturization of TEE probes designed for use in young infants weighing 3 to 3.5 kg or less has greatly enhanced the scope of TEE in the pediatric age group.2225 Successful TEE examinations have been reported in patients weighing as little as 2.3 kg.26,27 The role of TEE in pediatric cardiology is continuously evolving. Although the transthoracic window is adequate in most situations, TEE provides distinct advantages during cardiovascular surgery,22,24,2729 during video-assisted thoracoscopic procedures,30 for guidance of interventional catheter procedures,31,32 in the intensive care unit,22,33 for detection of intracardiac thrombi and vegetations,34 for assessment of prosthetic valves,35 in selected patients on mechanical assist device and extracorporeal membrane oxygenator,36 and in selected patients with poor transthoracic windows, such as adults with CHD.37 Whether to use intraoperative TEE selectively or routinely in patients with CHD deserves further study.21,38


Figure 107–6 Transesophageal imaging. A, Transverse plane: cross-sectional view at level 1a depicts the proximal ascending aorta (Ao), main pulmonary artery (MPA), and left and right pulmonary arteries (LPA and RPA). A rightward tilt of the transducer shows the RPA as it passes behind the superior vena cava (SVC) and ascending aorta (Asc Ao). To obtain a four-chamber view (level 3), the transducer is advanced in the esophagus with slight retroflexion of the scope. B, Level 2 is parallel to the transthoracic parasternal short-axis view. In level 2a, the atrial septum is imaged by a slight rightward tilt of the transducer. In level 2b, the aortic valve (AoV) is seen in cross section in the center of the image, and the left atrium (LA), right atrium (RA), tricuspid valve, right ventricular outflow (RV outflow), pulmonary valve (PV), and the proximal main pulmonary artery are seen. By advancing the transducer into the lower esophagus and anteflexing the scope, a cross-sectional view of the left ventricle (LV), mitral valve, and papillary muscles (PMs) is obtained (level 4). Note that image orientation is the same as in transthoracic echocardiography. C, Vertical (longitudinal) plane: the sweep begins at a plane that crosses the superior vena cava (SVC), left atrium (LA), right atrium (RA), and atrial septum (level 1). Next, a leftward tilt of the transducer shows an image parallel to the transthoracic parasternal long-axis view of the left atrium (LA), mitral valve, left ventricle (LV), left ventricular outflow tract, and proximal aorta (Ao) (level 2). Further leftward tilt of the transducer (level 3) shows the right ventricular outflow tract (RV outlet), pulmonary valve, and main pulmonary artery (MPA). The sweep continues leftward to show the leftward aspects of the left atrium, mitral valve, and left ventricle (level 4). Further leftward tilt depicts the left atrial appendage and the left pulmonary veins (not shown). Note that image orientation is the same as in transthoracic echocardiography.

Fetal Echocardiography

Examination of the human fetal cardiovascular system dates back to the late 1960s when continuous wave Doppler was used to record fetal heart rate. Although Kleinman and colleagues43 in the late 1970s had some success in detecting CHD in the fetus by M-mode echocardiography, it was not until high-resolution 2D imaging became available in the mid 1980s that accurate delineation of cardiovascular anatomy became clinically routine. Today, prenatal detection of CHD can be reliably diagnosed by 17 to 20 weeks of gestation by transabdominal imaging. Using the transvaginal window, the heart and great vessels can be imaged as early as late first trimester.44

Indications and Objectives

Although several studies have demonstrated a low detection rate of CHD by routine level I obstetric ultrasound,45 cost–benefit considerations preclude universal fetal echocardiographic screening by expert pediatric echocardiographers. Therefore, pregnancies at high risk for CHD are targeted. This approach increases the echocardiographic yield to approximately 30% when extracardiac anomalies had been detected, to approximately 60% when a level I scan had detected possible CHD, and to nearly 100% when a second opinion had been requested.46,47 The indications for fetal echocardiography are summarized in Box 107-1.

Description of Technique

Echocardiographic examination of the cardiovascular system in the fetus is based on the same segmental approach to diagnosis of CHD that is used in the newborn. The main difference between examination of the fetus and the newborn is that the operator has no control over fetal position and, consequently, over the views obtained. Once fetal position is ascertained and the spatial coordinates are determined, the examination continues according to the principles outlined in the previous sections. Given optimal acoustic windows and favorable fetal position, even the most complex cardiovascular anomalies can be detected (Fig. 107-7).48 Defects that remain difficult to diagnose in utero include secundum atrial septal defect, patent ductus arteriosus, small or moderate ventricular septal defect, coarctation of the aorta, and some valve and great vessel abnormalities.49 The distinction between normal patency of the foramen ovale and an atrial septal defect is usually not possible in the fetus. Similarly, it is not possible to predict whether the normally patent ductus arteriosus will close after birth. In utero diagnosis of aortic coarctation may be difficult because the typical isthmic narrowing may not become apparent until after ductal closure.49 However, the diagnosis may be suspected on the basis of abnormal morphology of the transverse aortic arch (elongation and hypoplasia), size discrepancy between the diameters of the aortic isthmus and patent ductus arteriosus, continuous flow profile in the aortic isthmus, and size discrepancy between the ventricles (right ventricle larger than left ventricle).50,51

Fetal Arrhythmia

One of the major reasons for referral to fetal echocardiography is arrhythmia. Most commonly, irregular fetal heart rate is caused by either conducted or blocked premature atrial contractions. These are usually benign and require follow-up only if they are frequent or associated with supraventricular tachycardia. Among the serious arrhythmias encountered in the fetus, supraventricular tachycardia and atrial flutter are the most common.46 Premature ventricular contractions and ventricular tachycardia are rare in the fetus, but both have been reported.46 Transplacental pharmacologic therapy is indicated in fetuses with incessant tachyarrhythmia, especially if signs of heart failure (e.g., enlargement of cardiac chambers, pericardial effusion, hydrops fetalis) are present. Complete heart block is the most common cause of prolonged bradycardia in the fetus.55 The distinction between complete heart block and sinus bradycardia is important, as the latter might indicate fetal distress. In fetuses with a structurally normal heart, maternal lupus erythematosus should be suspected. The most common structural CHDs associated with complete heart block are physiologically corrected transposition of the great arteries {S,L,L} and heterotaxy syndrome with polysplenia.

Fetal arrhythmias have been traditionally diagnosed by Doppler and M-mode echocardiography.46 Both methods rely on simultaneous recording of signals from the atria and ventricles. Fetal position, however, may not always allow optimal alignment of the M-mode cursor or the Doppler beam. A newer method, tissue Doppler imaging with a high frame rate, is a promising alternative to standard M-mode and Doppler echocardiography.56

Quantitative Analysis

Modern echocardiography allows accurate measurements of cardiovascular structures.1357 These measurements are helpful in deciding whether the size of a certain structure is within normal limits, and to quantify the degree of deviation from the expected norm.

Description of Technique

Measurements of linear dimensions (such as vessel diameter), cross-sectional areas (such as valve area), or volumetric dimensions (such as ventricular volume or mass) provide important quantitative information that can be used to assess the severity of the disease process and to predict its course and prognosis. Because the pediatric age group encompasses a wide range of body sizes and because the heart and great vessels grow considerably from birth to adulthood, measured dimensions must be adjusted to allow meaningful comparisons between patients of different ages and body sizes.58 For example, the mean value of the aortic valve anulus diameter is 0.74 cm in a 3.6-kg newborn (body surface area, 0.24 m2) and 1.95 cm in a 60-kg adolescent (body surface area, 1.66 m2). Indexing the aortic anulus diameter to body weight yields vastly different values (0.21 cm/kg in the newborn versus 0.032 cm/kg in the adolescent). Indexing the aortic valve diameter to the body surface area yields similarly unsatisfactory results (3.1 cm/m2 in the newborn versus 1.17 cm/m2 in the adolescent). Tanner and coworkers59 in 1949 and, more recently, Gutgesell and Rembold60 and Sluysmans and Colan58 noted that the growth of cardiac structures is not necessarily a linear function of the body surface area, the weight, the height, or the age. These findings are to be expected because the heart and great vessels grow much faster during the first 2 to 4 years of life compared with later childhood and adolescence. It was found that linear dimensions (such as diameters of valves and great vessels) should be indexed to the square root of the body surface area.5860 Returning to the preceding example of the aortic valve diameter, indexing the valve diameter to the square root of the body surface area yields 0.74/0.49 = 1.51 cm/m0.5 in the 3.6-kg newborn and 1.95/1.29 = 1.51 cm/m0.5 in the 60-kg adolescent (Fig. 107-8). Cross-sectional measurements (such as valve area) should be indexed to the body surface area.60 Left ventricular volume should be indexed to the body surface area raised to the power 1.28, and left ventricular mass should be indexed to the body surface area raised to the power 1.23.58

An alternative approach to comparing measurements between individuals is the use of z scores.1 The z score is a statistical expression of the position of a data point relative to the regression line of a data set. The z score is expressed as the number of standard deviations from the expected mean of a normal population. It is calculated as follows:


Returning to the preceding example of aortic valve diameter, the z score for a 0.74-cm aortic valve diameter in a newborn with a body surface area of 0.24 m2 is 0. That means that 0.74 cm is the mean value of aortic valve diameter in infants with that body size. In the adolescent with a body surface area of 1.66 m2, the z score of an aortic valve diameter of 1.95 cm is also 0. In other words, the same z scores in the newborn and the adolescent indicate that both values are in the same position relative to the regression line of normal values and thus are comparable. An aortic valve diameter of 0.53 cm in the newborn would have a z score of −2.0, which indicates that this value is 2 standard deviations below the expected mean. An aortic valve diameter of 0.96 cm in the same newborn would have a z score of +2.0, which indicates that this value is 2 standard deviations above the mean. Thus, expression of measurements as z scores allows comparison between patients, regardless of differences in body size.

An estimation of left ventricular volume is an important factor in determining the adequacy of chamber size in patients with left ventricular hypoplasia and in the evaluation of patients with volume overload. Numerous algorithms based on several geometric models have been developed for calculation of left ventricular volume, and they have been reviewed elsewhere.1,2 The biplane Simpson rule is considered among the most reliable methods for estimation of left ventricular volume. This method requires imaging of the left ventricle from two orthogonal views that share a common long axis: for example, the apical four- and two-chamber views (Fig. 107-9A). Left ventricular volume is calculated according to the following equation:


where ai is slice radius in the apical two-chamber view, bi is slice radius in the apical four-chamber view, L is left ventricular length, N is number of slices, and V is volume.

Another method for estimation of left ventricular volume is the biplane area-length method, where V = 5/6 × area × length (Fig. 107-9B). Experience with assessment of left and right ventricular volumes by 3D echocardiography suggests that this technique is potentially more accurate than 2D techniques.7,6165 Left ventricular myocardial volume can be measured from the 2D echocardiogram by subtracting the endocardial volume from the epicardial volume. Left ventricular mass is calculated by multiplying the resultant myocardial volume by the density of muscle (1.055 g/mL). Because it is not influenced by acoustic windows and is independent of chamber geometry, MRI provides an excellent alternative to echocardiography for measuring chamber volume and mass and is considered the reference standard to which other techniques are compared (see “Magnetic Resonance Imaging,” later).

Ventricular Function

Left ventricular function can be assessed at several levels. The heart may be viewed as a pump designed to maintain adequate flow to vital organs.66 This approach focuses on the external work performed by the heart, but it ignores the internal work and the functional state of the myocardium. Measuring cardiac output and systemic and pulmonary venous blood pressures can assess the pump function of the heart. It is known, however, that cardiac output and blood pressure can remain within the normal limits despite significant myocardial dysfunction.66

Ejection-phase indices of ventricular function, including shortening fraction, fractional area change, ejection fraction, velocity of circumferential fiber shortening (VCF), peak dP/dt, and systolic time intervals, measure global pump function.1366 Common to these indices is their dependence on loading conditions. These indices are unable to distinguish between the effects of altered loading conditions and abnormalities in myocardial contractility. Hence, abnormalities in preload and afterload can result in depressed shortening or ejection fractions, leading to the erroneous interpretation that myocardial contractility is depressed. On the other hand, left ventricular myocardial contractility may be depressed even in the presence of normal shortening or ejection fractions. The advantage of most of these indices is their relative simplicity and ease of acquisition. Load-independent assessment of left ventricular systolic function requires a more sophisticated analysis. The interested reader is referred to the relevant chapters in this book and to other reviews of this topic.66

These methods of ventricular function assessment rely in part on measurements of ventricular dimensions. An alternative approach is to evaluate myocardial motion and deformation.67 Several methods have been developed to measure myocardial velocities, strain, strain rate, displacement, and torsion (twist). One method uses Doppler to measure myocardial velocities (Doppler tissue imaging) and to calculate strain, strain rate, and other variables of tissue deformation (Fig. 107-10A).6870 Another method, called speckle tracking, analyzes the frame-to-frame changes in ultrasonic signal characteristics and uses that information to track the myocardium throughout the cardiac cycle (Fig. 107-10B).71 These methods are independent of ventricular geometry and provide information on systolic and diastolic function.7274 The reproducibility and the prognostic role of these techniques in patients with CHD are the subject of ongoing investigations.

< div class='tao-gold-member'>

Jul 31, 2016 | Posted by in CARDIAC SURGERY | Comments Off on Diagnostic Imaging: Echocardiography and Magnetic Resonance Imaging
Premium Wordpress Themes by UFO Themes