| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Perspective |
1 Department of Radiology, Hospital Calmette, Boulevard Pr. J. Leclerq, Lille
59037, France.
2 Present address: Department of Thoracic Imaging, The University of Texas M. D.
Anderson Cancer Center, Box 57, 1515 Holcombe Blvd., Houston, TX
77030-4095.
Received April 27, 2005; accepted after revision July 16, 2005.
Address correspondence to J. F. Bruzzi.
OBJECTIVE. MDCT systems with fast scanning capabilities can acquire images of the thorax with reduced cardiac motion artifacts, enabling improved evaluation of the heart and surrounding structures in the course of routine thoracic CT. This article describes the principles of including an evaluation of the heart in the course of a chest CT examination in terms of both examination technique and image interpretation. In addition, both the normal appearances and some of the most common abnormal appearances of the cardiac structures will be described.
CONCLUSION. Details concerning the cardiac structures can inform interpretation of thoracic CT studies and can influence the patient's clinical management. Both unenhanced and contrast-enhanced scans can detect significant cardiac disorders that may otherwise go undetected. In certain situations, a CT examination of the entire chest, complemented by cardiac gating, can provide a more dedicated analysis of the heart and coronary arteries, providing both morphologic and functional information.
Keywords: cardiac gating • cardiopulmonary imaging • chest • heart • MDCT • motion artifacts • thoracic CT
Thoracic CT is one of the most widely used imaging tests for the initial evaluation of thoracic disorders, but concurrent examination of the heart has traditionally been hampered by image degradation from cardiac motion artifacts. However, cardiac disorders can often complicate or coexist with extracardiac thoracic disease and may often remain unsuspected or underevaluated. Improvements in CT scanner technology—specifically, the development of subsecond MDCT scanners with high temporal and spatial resolution—have markedly shortened scanning times and now provide images of the heart during routine chest CT examinations that are much less degraded by cardiac motion artifacts and that allow detailed evaluation of the cardiac structures [1-9]. It should therefore be a goal of thoracic CT to screen patients for pericardial, cardiac, and coronary disorders in the course of routine clinical studies of the thorax requested by general physicians, respiratory physicians, or thoracic surgeons.
For more detailed cardiac evaluation, there also exists the capability of combining routine thoracic CT with supplementary ECG-gated studies of the heart and coronary arteries. These noninvasive cardiac gating techniques can provide important functional and morphologic information of the heart [10, 11] and can be applied in the course of conventional thoracic imaging for disease that may have a cardiac association. Details such as the presence of cardiomyopathy and ventricular aneurysms, the calculated ventricular ejection fraction, and cardiac output and ventricular mass, plus the analysis of myocardial contractility on CT cine images and assessment of myocardial perfusion can be relevant to physicians and surgeons and can usefully complement the routine assessment of a thoracic CT examination (Rémy J et al., presented at the 2004 annual meeting of the Journées Françaises de Radiologie).
These new possibilities for thoracic CT will inform the direction of future technologic developments, and they reflect the progressively increasing use of conventional mechanical CT in performing studies of the heart that were once feasible only with electron beam CT.
Technical Aspects: How to Include a Study of the Heart in a Thoracic CT Examination
Important information concerning the heart can be obtained on both unenhanced and contrast-enhanced CT, with or without cardiac gating. Table 1 provides a summary of suggested imaging techniques that may be applied to specific clinical questions.
|
Image Acquisition Without Contrast Enhancement
In certain situations, acquisition of a preliminary scan before the
administration of IV contrast medium can provide useful diagnostic
information. An unenhanced scan allows the natural density of the blood and
heart muscle to be appreciated and serves as a baseline for studying
myocardial enhancement patterns. In addition, the detection of calcification
in the heart valves, myocardium, pericardium, and coronary and pulmonary
arteries can be important for image interpretation. Obtaining an initial
unenhanced image acquisition with ECG gating can also allow formal calcium
scoring to be performed and can provide motion-free images of the pattern of
valvular calcification. Such information can be clinically useful,
particularly for risk stratification before anticipated surgery.
Image Acquisition With Contrast Enhancement but Without Cardiac Gating
In clinical practice, most CT scans of the thorax will be performed without
ECG gating. However, useful information of the heart and surrounding
structures can still be obtained, particularly with the latest generations of
MDCT scanners. In a patient with a pulse rate of 75 beats per minute, the
duration of diastole is 0.53 sec and that of systole is 0.28 sec. A CT scan
encompassing the entire cardiac region (12 cm craniocaudally) with a 16-MDCT
scanner, using a tube speed of 0.5 sec per rotation and a pitch of 1.5, will
include four to five cardiac cycles. It is therefore possible, during a single
acquisition, to obtain several image slices at different levels of the cardiac
chambers that are entirely in diastole, this phase being the period during
which the heart is the least mobile. Such images are relatively free of motion
artifacts and can provide information concerning ventricular wall thickness
and internal cardiac structures. The next generation of CT scanners, with tube
rotation times as fast as 0.33 sec per rotation, will allow cardiac imaging
with even greater spatial and temporal resolution.
Image Acquisition with Contrast Enhancement and with Cardiac Gating
Cardiac gating should no longer be regarded as an imaging tool limited to
studies of the heart. All of the anatomic structures surrounding the
heart—the mediastinum, the lung parenchyma, the pulmonary and systemic
vasculature, the pleural fissures, the tracheal bifurcation, and the
diaphragm—are subjected to cardiac pulsations and resultant motion
artifacts. The segmental and subsegmental pulmonary arteries of the
paracardiac regions alone account for 12 arteries that may be difficult to
analyze for reasons of cardiac motion artifacts. Cardiac gating may be useful
in selected cases (Table 1). In
particular, ECG gating is necessary for detailed depiction of cardiac and
coronary morphology and for analysis of cardiac function (Figs.
1A,
1B,
1C, and
1D). Cardiac gating is also
useful for improving evaluation of the pericardium and surrounding structures
and for vascular imaging of the thoracic aorta and pulmonary arteries in cases
in which it is desirable to minimize cardiogenic motion artifacts (e.g., in
cases of suspected aortic dissection or isolated segmental pulmonary
emboli).
|
|
|
|
Details concerning the acquisition of ECG-gated cardiac CT studies vary somewhat, depending on scanner specifications, and are described in several recent excellent articles [4, 5, 12, 13]. By correlating image acquisition with the electrical activity of the heart (ECG gating), retrospective image reconstruction can achieve temporal resolutions as long as 250 msec and as short as 92 msec, using a reconstruction algorithm of 180°. Nevertheless, current technology is still a long way from achieving completely motion-free cardiac imaging, which demands a temporal resolution between 30 and 50 msec [14]. In most patients undergoing cardiac CT, the success of ECG gating depends on the patient's having a regular cardiac rhythm at rates of less than 80 beats per minute, which allows a diastole of sufficient duration that can be temporally resolved from other phases of the cardiac cycle. In patients with relative tachycardias, it is often recommended that a ß-blocker be administered approximately 1 hr before CT, provided there are no contraindications to such medication (e.g., peripheral vascular disease, asthma, brittle diabetes).
The more recent generation of 16-MDCT scanners have tube speeds of 0.375 sec per rotation, permitting ECG-gated images of the entire thorax to be acquired in a single acquisition over 25 sec [7]. This allows the integration of ECG-gated studies of the heart and surrounding structures, free of motion artifacts, into CT examinations of the thorax in a single acquisition.
For those who do not have access to this more recent technology, a routine chest CT examination can incorporate a cardiac study by "coupling" a nongated scan with a second ECG-gated acquisition. However, the method of contrast administration and the sequence of which examination to perform first will depend on the clinical objective. We suggest the following general approach: For evaluation of the cardiac chambers, the myocardium, and surrounding paracardiac structures, an initial nongated contrast-enhanced CT scan of the entire thorax in the pulmonary arterial phase can be followed immediately by an ECG-gated acquisition of the heart using a supplementary bolus of contrast material. For a detailed analysis of the coronary arteries, cardiac valves, or pulmonary venous anatomy, the first acquisition should be ECG-gated scanning limited to the heart and performed during the phase of peak systemic arterial enhancement; it can be followed immediately by a second nongated acquisition of the entire thorax, complemented by a supplementary bolus of contrast medium. These two approaches maximize contrast enhancement of the structures of interest. Suggested imaging parameters for including an ECG-gated cardiac CT examination as a part of chest CT are outlined in Table 2.
|
Delayed Imaging
Patterns of myocardial enhancement after contrast medium injection have
been described that can help characterize regions of ischemia and infarction.
In the early phase of arterial enhancement, hypodense perfusion defects in the
myocardium that correspond to territories of the coronary arteries are a sign
of previous myocardial infarction
[15]. Visualization of
persistent areas of hypoattenuation on a second acquisition performed 5 min
later (a dual-phase scan) has been shown to correlate highly with fixed
infarcts depicted on subsequent technetium-99m sestamibi SPECT, whereas late
enhancement of these early defects is thought to represent areas of reversible
myocardial injury [16].
However, such techniques are still investigational and their clinical
application has yet to be determined.
Radiation Dose
The highly overlapping helical acquisition required for optimal
retrospective cardiac gating incurs a radiation dose of up to 10 mSv for a
single scan of the heart, which is greater than that of a conventional
coronary angiography study, or two to three times the annual background
radiation dose received by a person living in the United States
[17,
18]. Coupled scans that
incorporate a gated study of the heart in addition to a nongated study of the
entire thorax will expose the patient to even higher radiation doses, and the
expected benefit of such examinations should be carefully balanced against the
risks of high radiation exposure. However, substantial radiation dose
reductions of 30-50% can be achieved by in-line dose modulation so that the
tube output is reduced during systole, but this requires that the heart rate
be reasonably regular without significant tachycardia
[19,
20]. Nevertheless, concerns
persist as to these high radiation doses, and ECG-gated studies should always
be tailored to specific clinical indications. However, in certain situations,
a detailed evaluation of the heart can inform the interpretation of the
thoracic CT images and offer important supplementary information to the
clinician.
Image Reconstruction
In ECG-gated cardiac CT studies, the highly overlapping data that are acquired are normally reconstructed in different phases of the cardiac cycle (up to 20 phases, or at each 5% of the R-R interval), generating an extremely large number of images. To minimize the difficulties involved in generating, storing, displaying, and interpreting all of this information, image reconstruction should be tailored according to the clinical question of interest. An initial image data set will normally be reconstructed in the phase of peak diastole, which is the moment of maximum relaxation of the myocardium and associated with the fewest motion artifacts. Further image reconstruction in other phases of the cardiac cycle can be performed if information concerning cardiac function is desired, such as calculation of ejection fraction (Figs. 1A, 1B, 1C, and 1D), global cardiac output, or myocardial mass, or for the creation of cine images for analysis of myocardial contraction. Images are generally reconstructed in multiple planes for optimal depiction of the different regions of the heart and are the same as those used in cardiac MRI [21]. A thorough knowledge of cardiac anatomy is the basis of interpretation of cardiac CT studies [22].
|
|
|
The emphasis in this discussion will be on the right heart chambers, which are those most often neglected in cardiac imaging. Many of the following findings can be observed both with and without cardiac gating. In many cases, important morphologic and, to a lesser extent, functional information can be gleaned on the basis of visual cues alone.
Ventricles
The ventricular dimensions are best appreciated with their small axes
perpendicular to the interventricular septum, the small axis of the right
ventricle being roughly parallel to that of the left ventricle (Figs.
1A,
1B,
1C, and
1D). On axial images, the right
ventricle has a triangular appearance, whereas the left ventricle is more
ovoid. The maximum internal diameter of the right ventricle in its small axis
should be equivalent to that of the left ventricle.
Standardized measurements of normal myocardial thickness and intracardiac dimensions have been described for echocardiography and MRI [23-25], although they have not yet been established for CT. The normal myocardial thickness of the right ventricle is 3-4 mm, approximately three times thinner than that of the left ventricle [26]. Thickening of the wall of the right ventricle is a sign of elevated right ventricular outflow pressure.
The interventricular septum normally has an axis inclined 45° to the coronal and sagittal planes, and it should have a gentle convex curvature toward the right ventricle. It does not normally measure greater than 13 mm in thickness. The convexity of the interventricular septum is a marker of the function of the right ventricle [27]. There exists a ventricular interdependence so that, when the pericardium is normal, the left ventricle can contribute from one fifth to two thirds of the function of the right ventricle [28]. Similarly, a left shift of the interventricular septum resulting from dilatation of the right ventricle can impede filling of the left ventricle during diastole.
In the right ventricle are two functionally distinct chambers: the trabeculated inflow tract that holds most of the ventricular blood during diastole, and the smoother outflow tract that serves as the base of the pulmonary artery. In the inflow tract, the anterior and posterior trabecular muscles, and the smaller medial papillary muscle, are connected to the tricuspid valve leaflets via the chordae tendineae (Fig. 2A). The moderator band is a prominent muscular extension conveying the electrical apparatus of the right bundle of His from the interventricular septum to the anterolateral aspect of the right ventricle in the region of the ventricular apex, where it inserts at the base of the anterior papillary muscle (Fig. 2B) and separates the trabeculated inflow tract of the right ventricle from the smoother outflow tract [29]. Similarly, anterior and posterior papillary muscles can be visualized in the left ventricle connecting to the leaflets of the mitral valve via fine chordae tendineae (Fig. 3).
|
|
|
Atria
The obliquity of the interatrial septum on axial images is similar to that
of the interventricular septum—that is, about 45° to the horizontal.
The anteromedial portion of this septum is separated from its posterolateral
portion by the fossa ovalis. The multiple trabeculations and papillary muscles
of the endoluminal surfaces of the atria and ventricles can be quite bulky and
should not be mistaken for endocavitary thrombus or tumor. In particular, the
crista terminalis should be recognized as the muscular prominence on the
posterolateral aspect of the right atrium extending from the orifice of the
superior vena cava to that of the inferior vena cava
[29]
(Fig. 4). Chiari's network is a
lax tissue representing remnants of the eustachian valve; the network extends
from the inferior part of the crista terminalis to the base of the interatrial
septum and can simulate interatrial thrombus
[34].
Fat in the interatrial septum can be particularly abundant. Fat in the anteromedial part of the interatrial septum, situated anterior to the fossa ovalis, is generally more prominent than fat located more posteriorly. In both locations, the normal interatrial septum should measure no more than 10 mm in thickness [35]. When the amount of interatrial fat causes the septum to be thicker than 10 mm, the term "lipomatous hypertrophy of the interatrial septum" is used, signaling a possible association with cardiac arrhythmia [36] (Fig. 5). Lipomatous hypertrophy of the interatrial septum usually spares the fossa ovalis. The septum within the fossa ovalis is often difficult to visualize on CT images because of its thinness and because of streak artifacts from the bolus of contrast medium in the right atrium. In cases of right heart hypertension, the fossa ovalis can project into the left atrium, simulating a patent foramen ovale (Fig. 6), which is present in 25-30% of the population [37]. In such cases, the only way to show an actual shunt is to visualize the jet of contrast medium passing from the right atrium into the left atrium. Other, indirect signs of a patent foramen ovale include early contrast opacification of the left heart chambers and aorta before contrast arrival in the right side of the heart.
|
|
|
Pericardium
The pericardium is readily identified in most patients, being outlined by
both epicardial and pericardial fat. A thorough familiarity should be obtained
with the many varied appearances of the pericardial recesses, which can
simulate mediastinal adenopathy in the right paratracheal region, in the
aortopulmonary window, in the subcarinal region, and around the insertion of
the pulmonary veins [39,
40]. Pericardial effusion or
thickening may be abnormal and may be associated with restrictive
pericarditis. Calcification of the pericardium may indicate restrictive
pericarditis from previous tuberculosis or radiation therapy, whereas
calcification of the myocardium itself can signal a previous myocardial
infarction
[41-43]
(Fig. 8).
Functional Evaluation of the Heart
Measurement of ventricular volumes in systole and in diastole is performed on images obtained using cardiac gating and reconstructed along the small axis of the heart. For the right ventricle, this requires that the tricuspid and pulmonary valves be identified. Measurement of ventricular volumes allows calculation of the right ventricular ejection fraction and is based on the anatomic changes between systole and diastole (Figs. 1A, 1B, 1C, and 1D). Such techniques have been shown to correlate well with similar methods of calculating ejection fraction on MRI [44], although standards of normal and abnormal ejection fractions have not yet been established using CT. The right ventricular ejection fraction can also be obtained on echocardiography, radionuclide ventriculography, or more invasive techniques such as contrast ventriculography or thermodilution [45]. Echocardiography is a qualitative technique based on geometric assumptions and is poorly adapted to the crescent-shaped and bicompartmental form of the right ventricle [46]. In the context of thromboembolic disease, in which the development of pulmonary hypertension accompanies changes in ventricular morphology, a quantitative approach that takes into better account the complex relationship between function and anatomy of the ventricular chambers is required, for which CT is well suited.
Other functional information can be obtained by specific CT techniques:
Visual analysis of the progress of the bolus of contrast medium through the cardiac chambers and great vessels can provide clues to intracavitary abnormalities. During an acquisition in the craniocaudal direction, the left ventricle and aorta should be well opacified by contrast medium only toward the end of scanning. The existence of a left-to-right shunt by way of a patent foramen ovale can disturb the normal physiologic passage of contrast medium and cause the premature appearance of contrast medium in the left ventricle and aorta.
The cardiopulmonary transit interval of a bolus of contrast medium is the time between the peak of contrast opacification in the main pulmonary artery and the peak in the ascending thoracic aorta [47]. Calculation of the cardiopulmonary transit interval, which is performed routinely in cardiac MRI studies, can be used with equal validity in cardiac CT. The cardiopulmonary transit time is directly correlated to the diastolic and systolic intracavitary volumes of the left ventricle. According to calculations on MRI, the mean transit time is 7.2 ± 1.2 sec in the normal state and rises to 10.4 ± 3.3 sec in the presence of cardiac decompensation [48].
The cardiac output is calculated by certain software applications on the basis of semiautomated tracing of the endocavitary contours of the left ventricle in the phases of maximum systole and maximum diastole, which allows an estimation of the ejection volume. The cardiac output is calculated by multiplying the measured volume ejection by the heart rate during image acquisition. It can also be calculated from the time-density curve of a test bolus of 20 mL of contrast medium followed by a 30-mL saline push [47], a method adapted from the principles of the dilution of a color indicator. The results of these two techniques are closely correlated.
Calculation of the pulmonary vascular resistance can be derived from measures of the cardiac output and the cardiopulmonary transit time [49].
Summary
CT of the thorax offers the opportunity for an assessment of the heart for cardiac disease or anomalies that may otherwise pass unsuspected and undetected. With the increasing use of MDCT scanners and improved temporal resolution, images of the heart can be obtained that are relatively motion-free. For more detailed analysis of cardiac structure and function, an ECG-gated acquisition can usefully complement a routine chest CT examination in many situations. In either case, correct analysis of images of the heart and surrounding structures depends on a thorough understanding of cardiac anatomy and on recognition of the normal and abnormal appearances of the heart that are commonly observed on CT.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J. F. Bruzzi, M. Remy-Jardin, D. Delhaye, A. Teisseire, C. Khalil, and J. Remy When, Why, and How to Examine the Heart During Thoracic CT: Part 2, Clinical Applications Am. J. Roentgenol., February 1, 2006; 186(2): 333 - 341. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
