Abstract
Echocardiography maintains a central role in cardiovascular assessment despite recent advances in multimodality imaging. Some of the advantages of echocardiography include portability, rapid availability, low cost, safety and excellent temporal resolution. This in combination with continued technological development is the reason why cardiac ultrasound is the imaging tool of first choice in almost all conditions bar the assessment of flow limiting coronary stenosis. Cardiac magnetic resonance imaging, coronary computed tomographic angiography, single photon emission computed tomography and positron emission tomography each have unique advantages. The strengths and weaknesses of the different imaging modalities are discussed in this chapter and their utility outlined in conjunction with echocardiography. The synergistic relationship between echocardiography and multimodality imaging is emphasized. Seven main topics are addressed when considering echocardiography in the context of other imaging modalities and include volumes and function, morphology, tissue characterization, coronary disease, cardiac masses, pericardial disease and diseases of the aorta.
Keywords
cardiac computed tomography, cardiac magnetic resonance imaging, echocardiography, multimodality imaging, nuclear cardiology
Introduction
Despite advances in imaging technology over the past 20 years, echocardiography has maintained its central role in cardiovascular medicine. Although some of this relates to the fact that echocardiographic technology is also progressing steadily, the primary reason is by virtue of the unique advantages of echocardiography including portability, rapid availability, safety, and excellent temporal resolution. As a result, echocardiography is generally the first tool implemented for the evaluation of a wide variety of clinical indications. Its superior temporal resolution permits better assessment of small or thin mobile structures as well as the physiologic and hemodynamic consequences of disease. Although echocardiography is effectively a real-time imaging modality regarding temporal resolution, cardiac magnetic resonance (CMR) imaging comes in second place, with nuclear imaging and coronary computed tomographic angiography (CCTA) lagging further behind ( Table 48.1 ). On the other hand, the spatial resolution of CCTA and CMR is superior compared with echocardiography. Although spatial resolution is improved with positron emission tomography (PET) compared with single-photon emission computed tomography (SPECT), this imaging characteristic is the Achilles heel of nuclear imaging. Along with ionizing radiation, these properties of the four main imaging modalities are invariably taken into account when considering further evaluation with noninvasive cardiac imaging techniques. Careful consideration should also be given to the cost effectiveness of a diagnostic pathway.
| Transthoracic Echocardiography | Transesophageal Echocardiography | Computed Tomography | Magnetic Resonance | Spect And Pet | |
|---|---|---|---|---|---|
| Availability | Readily available Portable/bedside if required | Available in most centers but operator dependent | Cardiac CTA only available in centers of expertise | CMR only available in centers of expertise | SPECT usually available PET only available in centers of expertise |
| Cost | Low | Low to intermediate | Intermediate | High | Intermediate to high |
| Safety | Safe | Procedural complications are relatively rare | Ionizing radiation (1–14 mSv) Iodinated contrast usually C/I if eGFR <30 mls/min | No radiation Gadolinium based contrast C/I if eGFR <30 mls/min C/I if patient has a device, for example, PPMs a , ICDs a , LVADs | Ionizing radiation (3–21 mSv) |
| Arrhythmia | Impairs quality of 3D imaging (stitch artifact) | Impairs quality of 3D imaging (stitch artifact) | Impairs gating resulting in suboptimal images | Impairs gating resulting in suboptimal images | Impairs gated images |
| Patient-related factors limiting image quality | Limited windows Narrow field of view Limited by obesity, COPD, and postoperative setting | Esophageal pathology Sedation usually required | Hemodynamically stable only Poor breath-holder | Hemodynamically stable only Claustrophobia Poor breath-holder | Hemodynamically stable only |
| Temporal resolution | Almost real-time Long loop acquisitions possible | Almost real-time Long loop acquisitions possible | 160–200 ms Retrospective acquisition required for full cardiac cycle | 30–50 ms with SSFP | 8–16 frames per cardiac cycle |
| Spatial resolution | Dependent on US frequency and depth | Better than transthoracic (increased US frequency possible) | 0.5 mm with later generation CT scanners | 5–8 mm 0.9–1.2 mm when ECG and respiratory gated (3D acquisition) | 9–10 mm A-SPECT or 4–5 mm D-SPECT 4–5 mm PET |
| Myocardial perfusion imaging | Contrast echo with microbubbles—promising but not FDA approved | Not applicable | Fair—not yet mainstream. Improves diagnostic accuracy of CT | Good | Excellent PET > SPECT |
| Pericardial assessment | Best modality for evaluating hemodynamic effect | Reasonable for anatomic assessment | Excellent for anatomy (including calcification) | Excellent for both anatomy and hemodynamic effect | Not applicable |
| Tissue characterization | Limited Contrast enhanced | Limited | Good Hounsfield unit | Excellent | Good Molecular imaging techniques for scar, inflammation, etc. |
| Extracardiac structures | Limited | Limited | Excellent | Excellent | Limited unless performed with CT attenuation correction |
a Note is made of MRI-safe PPMs and the fact that some centers perform CMR on patients with ICDs.
It is rare for a patient to undergo a CMR, CCTA, SPECT, or PET without first having an echocardiogram. Multimodality imaging and echocardiography have a complementary or synergistic relationship. It is always extremely useful to review the prior imaging of a patient when considering the next best test or when protocoling and subsequently analyzing the new study. To address the role of multimodality imaging in the context of echocardiography, seven main topics are discussed as outlined in Fig. 48.1 .
Chamber Quantification: Volumes and Function
Echocardiography is the mainstay, although not the gold standard, for evaluating chamber size and function through linear measurement, and volume quantification. End-diastolic and end-systolic dimensions and volumes are the most commonly used parameters to describe left ventricular cavity size and function. Volumes derived from linear measurements are confined to the assumption that the left ventricle (LV) is a fixed geometric shape, such as a prolate ellipsoid. As a result, the Recommendations for Cardiac Chamber Quantification by Echocardiography in Adults outline that the Teichholz and Quinone methods for calculating LV volumes are no longer recommended for clinical use. The recommended method for two-dimensional echocardiographic (2DE) volume calculation is the biplane method of disks summation or modified Simpson’s rule. Volume is generated by taking an average of the apical four- and two-chamber views, and the measurements are standardized by indexing to body surface area. Foreshortening of the ventricle and wall motion abnormalities can cause erroneous results. Poor endocardial definition is improved by administration of contrast agents. Although contrast-enhanced images provide larger volumes than unenhanced images, these measurements are closer to those found with cardiac magnetic resonance imaging (MRI). Three-dimensional echocardiographic (3DE) volume measurement is accurate and reproducible and eliminates the error introduced by geometric assumptions in 2DE. It should also be remembered that volume measurements from different imaging modalities should not be used interchangeably. For instance, CT can overestimate and echocardiography underestimate right ventricular (RV) volume when compared with the gold standard CMR. Although accurate and comparable to CMR, 3D-transthoracic echocardiography (TTE) and 3D-transesophageal echocardiography (TEE) tend to underestimate volume and function.
Accurate estimation of LV systolic function is of particular importance in patients under consideration for an implantable cardioverter defibrillator or resynchronization therapy, for those whose profession is dependent on a certain cutoff for LV ejection fraction (LVEF), and in patients undergoing chemotherapy with cardiotoxic agents. Deformation imaging by echocardiography using strain or two-dimensional (2D) speckle tracking permits accurate assessment of global and regional LV systolic function. Strain analysis increases the sensitivity for detecting subclinical myocardial dysfunction in cardiomyopathies before there is an overt drop in LVEF. Multimodality imaging is required when there is some doubt as to the exact LVEF. In this setting, multigated acquisition (MUGA) nuclear imaging or CMR are usually performed and are considered to be the most accurate and reproducible methods for the assessment of LV systolic function. Retrospectively gated CCTA also provides a very accurate evaluation of LVEF but is generally reserved for other indications.
Assessment of RV size and function is challenging due to the complex geometry of the RV chamber. On TTE, RV size is often estimated by visual assessment or “eyeballing.” Various surrogates are used to assess size and function, each with their own limitations. 3DE allows volumetric assessment and complements 2D measurements. Although 3DE is accurate, it is time consuming and tends to underestimate volumes compared with CMR. Few laboratories measure RV size and function by 3DE. CMR is the gold standard for the assessment of RV size and function. The usefulness of the echocardiographic “eyeball” method to estimate RV size and systolic function in patients with right heart disease is limited when compared with CMR, specifically with regard to interobserver variability between echocardiographers. Cardiac CT provides accurate and reproducible RV volume measurements when compared with CMR and, like 3DE, is an alternative for patients who have a contraindication to CMR.
The “overloading” effect of chronic regurgitant valve lesions, as well as intracardiac and extracardiac shunts, is best measured by the degree of ventricular dilatation. The accurate assessment of the severity of mitral regurgitation (MR) and aortic insufficiency (AI) by echocardiography is difficult when the lesions are in the moderate to severe range. Although the vena contracta, effective regurgitant orifice area, and flow reversal (pulmonary vein in MR and holodiastolic flow reversal in the aorta in AI) are all useful measures of the severity of valvular dysfunction, regurgitant volume, regurgitant fraction, and LV dilatation are key when determining the need for surgical or transcatheter intervention. Echocardiography is an excellent tool for assessing and monitoring valvular heart disease and carries a class 1B indication for AI and MR. CMR is indicated in patients with moderate or severe AI (stages B, C, and D) and suboptimal echo images for the assessment of LV systolic and diastolic volumes, function, and measurement of AI severity. Like echo, CMR has a class 1B recommendation for the evaluation of AI and is a useful screening tool for associated aortopathies. Aortic magnetic resonance angiography (MRA) or CTA is also indicated in patients with a bicuspid AV when morphology of the aortic sinuses, sinotubular junction, and thoracic aorta cannot be assessed accurately or fully by echo (class 1C). Serial assessment of size and morphology can be achieved with any one of the three modalities (class 1C). CMR is also indicated in patients with chronic primary MR to assess LV and RV volumes, function, and MR severity when these issues are not satisfactorily addressed by TTE (class 1B). There is also evidence to suggest that CMR should be considered when MR severity as assessed by TTE is influencing important clinical decisions, such as the decision to undergo mitral valve surgery ( Figs. 48.2 and 48.3 ; 
). TEE is indicated for the evaluation of chronic primary MR in whom noninvasive imaging provides nondiagnostic information about the severity of MR, mechanism of MR, and/or status of LV function (class 1C).
) indicated by the green arrow . An eccentric anteriorly directed jet of mitral regurgitation is demonstrated in A (see 
) and B (TTE A4Ch) indicated by the white arrow . Color M-mode is consistent with holosystolic mitral regurgitation ( white arrow , F). The mitral regurgitant volume using the PISA formula was 38 mL (D and E, PISA radius indicated by the white arrow ). Systolic flow across the ASD is seen in C (TTE A4Ch, see 
) indicated by the orange arrow . CMR images are shown in Fig. 48.2B . A4Ch, Apical four-chamber view; ASD, atrial septal defect; CMR, cardiac magnetic resonance imaging; PISA, proximal isovelocity surface area; PSLA, parasternal long axis; TTE, transthoracic echocardiogram. 
) indicated by the green arrow . The eccentric anteriorly directed jet of mitral regurgitation is demonstrated in B (3Ch FSPGR) indicated by the white arrow . Systolic flow across the ASD is seen in C (4Ch SSFP, see 
) and indicated by the orange arrow . Left and right ventricular end diastolic and end systolic volumes were obtained by tracing the endocardium of every slice on the CMR short-axis stack (D, SSFP and see 
) providing LVSV and RVSV. Pulmonary artery and aorta forward flow volumes were also assessed using phase contrast sequences (flow curves shown in E). Qp:Qs was 1.4:1 and mitral regurgitant volume was 91 mL. This case illustrates how mitral regurgitant volumes can be underestimated on TTE when a significant ASD is present and the utility of CMR in such cases. 4Ch, Four-chamber view; 3Ch, three-chamber view; ASD, atrial septal defect; CMR, cardiac magnetic resonance imaging; FSPGR, fast spoiled gradient echo; LVSV, left ventricular stroke volume; PISA, proximal isovelocity surface area; RVSV, right ventricular stroke volume; SSFP, steady-state free precession imaging; TTE , transthoracic echocardiogram. Unexpected chamber dilatation raises the question of a cardiovascular shunt and is often a finding noticed on TTE. The right atrium and right ventricle are usually dilated when there is a significant shunt proximal to the tricuspid valve, whereas a shunt distal to the tricuspid valve results in LV dilatation (e.g., ventricular septal defect or patent ductus arteriosus). There are other clues on TTE as to the presence of a shunt; for instance, RV function is usually preserved, and there are increased velocities across the pulmonary valve in shunts proximal to the tricuspid valve. TEE can be performed to localize and define morphology of the shunt. Qp:Qs can also be evaluated during TEE and is defined as (CSA RVOT × VTI RVOT )/(CSA LVOT × VTI LVOT ). A patent foramen ovale is best picked up on echocardiography compared with other imaging modalities. Although echocardiography and angiography have traditionally been the primary tools used for the evaluation of cardiac shunts, CCTA and particularly CMR are proving to be very useful in this setting.
CCTA is an excellent modality for demonstrating structural heart disease including septal defects, anomalous pulmonary venous return, and arterial and venous anatomy but is limited in terms of functional analysis. Volumetric analysis can be performed with CT but is hampered by the presence of regurgitant valve lesions and higher doses of radiation with the necessity for retrospective gating. CMR has emerged as an accurate noninvasive alternative for characterization of anatomy and assessment of function. Precise shunt quantification is achieved with CMR using both volumetric cine imaging and phase-contrast cine imaging (see Fig. 48.3 ; see 
). Phase contrast techniques derive contrast between flowing blood and stationary tissue by manipulating the phase of the magnetization. This method provides accurate measurements of volumes and velocities in the absence of significant turbulence and is the most accurate of the noninvasive modalities.
Morphology
LV mass is a prognostically important parameter and should be reported in imaging studies where possible, particularly in patients with hypertension. There are different methods to measure LV mass on TTE, including M-mode and 2DE linear equations, and other 2D-based formulae, such as the truncated ellipsoid and the area-length method. For the linear method, mass is calculated using the unidimensional Devereux formula indexed to body surface area and relies on a normally shaped LV. 3DE has an advantage over 2DE in that potential errors from geometric assumptions are removed. In fact, the estimation of LV mass by 3DE has similar accuracy to the gold standard CMR. Although generally not performed, LV mass may also be assessed by CT with reasonable accuracy. Estimation of mass is less accurate with nuclear techniques.
Abnormal morphology in the setting of congenital or acquired heart disease is usually identified on TTE initially. Pattern recognition allows the experienced sonographer and reader to infer a likely cause or differential diagnosis. Dilated chambers as described previously provide a hint as to the abnormality, but it is anatomic imaging in the form of 3D volumes that permits precise localization of the defect or altered anatomy. Volumetric imaging in 3DE, CCTA, and CMR permits multiplanar reconstruction of congenital abnormalities. Indeed, multimodality imaging is paramount in terms of diagnosis, decision to intervene, and suitability for different therapies, as well as monitoring and follow-up. Congenital heart disease (CHD) is something that adult cardiologists are now seeing more often as infants born with CHD survive into adulthood and require lifelong follow-up and care. Although echocardiography is routinely used in all patients, complementary multimodality imaging is frequently required particularly in postsurgical patients or those with complex CHD. Three-dimensional volume sets with both CT and CMR allow comprehensive assessment of cardiac anatomy, the aorta, pulmonary arteries, and venous return. Quantification of ventricular volumes and function can be achieved with 3DE, CT, and CMR, as discussed. Although CT has superior spatial resolution, CMR has the benefit of complex flow measurements for shunt evaluation.
Anomalies of the coronary arteries are rare, with an incidence of 0.2%–1%. Patients presenting with symptoms resulting from coronary artery anomalies are usually younger and can present with symptoms relating to episodic myocardial ischemia such as chest pain, syncope, ventricular arrhythmia, or sudden cardiac death. In patients with suspected anomalous coronary arteries, echocardiography may establish the diagnosis, but the predictive value remains controversial. Pediatric cardiologists are better versed in looking for anomalous origins of the coronary arteries by TTE compared with adult cardiologists. A particular strength of CCTA is the depiction of coronary artery anatomy including the ostia and course of the coronary vessels (refer to Fig. 48.12 , later). The appropriate use criteria ascribe a score of 9 to CCTA for the evaluation of anomalous coronary arteries. CMR is also useful in delineating the proximal course of the coronary arteries and is often considered in younger patients when avoidance of ionizing radiation is preferred.
Another instance in which an initial evaluation with echocardiography leads to further multimodality imaging is acquired abnormal chamber morphology. Examples include excessive hypertrophy/hypertrophic cardiomyopathy, dilated and restrictive cardiomyopathies of uncertain etiology, LV noncompaction, endomyocardial fibrosis, and occasionally Takotsubo cardiomyopathy ( Fig. 48.4 ; 
) to name but a few. CMR is usually recommended when a new diagnosis is made or suggested by TTE. As well as excellent anatomic imaging, CMR offers the additional benefit of tissue characterization, which may be pathognomonic in certain cardiomyopathies. After the diagnosis has been made by complementary imaging techniques, surveillance and follow-up are usually accomplished with TTE.
). Left heart catheterization reveals the classical “octopus pot” appearance in systole (F, orange arrows: akinesis of all mid and apical segments, white arrows: hyperdynamic basal function). There was nonobstructive coronary disease (G and H showing left and right coronary arteries respectively). CMR was performed to rule out myocarditis. 4 and 2Ch SSFP demonstrate the previously mentioned wall motion abnormalities (I and J, see 
). Fat-suppressed T2-weighted imaging suggests inflammation in the mid (green arrows) and apical segments (K). There was no hyperenhancement on late gadolinium enhancement imaging consistent with the absence of infarction or infiltration and a finding consistent with a diagnosis of Takotsubo cardiomyopathy (L). An apical thrombus (red arrow) was discovered on post gadolinium long TI time sequences (M). Repeat TTE on admission day 5 demonstrated complete recovery of left ventricular systolic function (see 
). The apical thrombus persists and is identified using contrast-enhanced TTE (see 
). A4Ch, Apical four-chamber view; A2Ch, apical two-chamber view; CMR, cardiac magnetic resonance imaging; SSFP, steady-state free precession imaging; TTE, transthoracic echocardiogram. Tissue Characterization
Characterization of myocardial tissue provides virtual in vivo histology in patients with cardiomyopathy and unexplained increased LV mass. There are a variety of potential causes of excessive mass beyond hypertension and aortic stenosis, which may have important implications in terms of treatment and prognosis. Further evaluation should be considered when the degree of LV hypertrophy or wall thickening is out of keeping with clinical picture. Options for the evaluation of myocardial tissue characterization by TTE are limited and include measures of tissue reflection (a surrogate for tissue density) and dynamic changes as a result of alterations in myocardial ultrastructure. Ultrasonic scatter from small reflectors also known as integrated backscatter is the only available echo tool to evaluate tissue density. Collagen within the tissue results in scatter and attenuation and is the primary determinant; however, feasibility with this method is limited. On the other hand, tissue characterization is a major strength of CMR and to less extent cardiac PET.
Inflammation, infarction, and infiltration are demonstrated on CMR primarily with late gadolinium enhancement (LGE) but also with other specialized sequences permitting characterization of pathologic tissue without the need for contrast. Long T2 relaxation times of water-specific contrast results in high signal intensity of edematous tissue. Harnessing this property in combination with black blood imaging allows detection of acute myocyte swelling and interstitial fluid accumulation. This technique is used to identify acute inflammation in myocarditis, acute myocardial infarction, stress cardiomyopathy (Takotsubo), and transplant rejection and will be somewhat refined with the advent of T2 mapping. On the other hand, T1-weighted sequences result in a high signal intensity of fatty tissue. These sequences are useful in the assessment of the pericardium as epicardial, and pericardial fat layers provide an excellent contrast for margins of the visceral and parietal pericardium respectively. Fatty infiltration in the context of arrhythmogenic RV cardiomyopathy or cardiac tumors is also identified with this technique. Iron accumulation in the myocardium is evaluated by CMR in the form of a unique myocardial T2 star map ( Fig. 48.5 and 
). This method has been validated with in vivo histology and allows direct detection and quantification of myocardial iron in vivo. There are three ways in which myocardium can be characterized with gadolinium contrast agents. As the contrast is administered, rest or stress myocardial first pass perfusion is observed with dynamic imaging. Early gadolinium enhancement occurs within 1–3 minutes of injection and LGE is observed 5–20 minutes post administration. Gadolinium is a heavy metal that accumulates in expanded extracellular space and when present LGE represents inflammation (edema), fibrosis, or infiltration. The pattern of enhancement is crucial, with LGE extending from the endocardium across the myocardium (subendocardial LGE) usually indicating varying degrees of myocardial infarction. The specificity of the LGE pattern in ischemic cardiomyopathy has been confirmed in numerous studies in which coronary angiography and CMR were performed in patients with systolic dysfunction of unknown etiology. Mid-wall or subepicardial LGE is found in myocarditis as well as various cardiomyopathies. The segmental location, distribution (focal or patchy) and extent of LGE often holds a clue as to the underlying pathologic process. Occasionally extensive systemic amyloidosis involving the heart results in a classical dilemma where the myocardium cannot be adequately “nulled” due to altered gadolinium kinetics relating to sequestration of gadolinium by other organs infiltrated with amyloid plaques ( Figs. 48.6 and 48.7 ; 
). T1 mapping will further differentiate the underlying cause of various cardiomyopathies as well as permit earlier detection of myocardial interstitial fibrosis.
