Current Clinical Applications of Cardiac Computed Tomography
- First Online:
- Cite this article as:
- Achenbach, S. J. of Cardiovasc. Trans. Res. (2011) 4: 449. doi:10.1007/s12265-011-9278-y
- 95 Views
Multi-detector row CT allows high-resolution anatomic and functional visualization of the heart. The main current clinical application is non-invasive coronary angiography after intravenous injection of contrast agent. Coronary CT angiography permits the detection of coronary artery stenoses with relatively high accuracy and especially high negative predictive value. It is therefore considered clinically useful to rule out disease and avoid invasive coronary angiography in patients with low to intermediate likelihood of disease and especially with equivocal stress test results. Of lesser clinical relevance, albeit indicated in certain patient subgroups, is the use of cardiac computed tomography for coronary calcium quantification in the context of risk stratification. Finally, the analysis of cardiac morphology and function, including the assessment of valvular disease, is possible by CT. For most of these applications, CT is only indicated if more standard techniques such as echocardiography or cardiac magnetic resonance fail. Correct patient selection as well as sufficient experience and expertise are prerequisites for the beneficial clinical application of coronary CT angiography.
KeywordsComputed tomographyHeartCoronary artery diseaseCoronary calcificationDiagnosis
The technical evolution of computed tomography has been rapid and continues up to today. While the first computed tomography systems had a spatial and temporal resolution which were substantially lower than needed for cardiac imaging, the introduction of so-called multi-detector row computed tomography systems around the year 2000 helped overcome a number of obstacles at the same time: The systems provided thin (sub-millimeter) slice collimation and therefore were capable of resolving the small structures of the heart and coronary arteries. At the same time, multiple cross sections could be acquired simultaneously (initially 4 and by now up to 320) so that in spite of the use of thin slices, the volume of the heart could be covered in a reasonably short time (well within one breathhold). A further aspect was the development of ECG-gated image acquisition and reconstruction. ECG gating is necessary to ensure that all data of one image set are obtained at the identical cardiac phase. Finally, increasing gantry rotation speed provided the temporal resolution necessary to “freeze” cardiac motion. In fact, the need for ECG gating (and, consequently, vulnerability to arrhythmias) as well as a somewhat limited temporal resolution remain the major problems of cardiac computed tomography up to today.
Until very recently, 64-slice computed tomography systems constituted the “state of the art” for cardiac CT imaging. However, technical evolution continues at a rapid pace. On one hand, the number of detector rows has been increased to as many as 320 (which under certain circumstances permits imaging of the entire heart in one single heart beat). On the other hand, “dual source CT,” which combines two X-ray tubes and two detector arrays in a single gantry, improves temporal resolution by a factor of 2 and thus reduces image artifacts.
The constant improvement of CT technology, the increasing penetration of high-end CT systems capable of cardiac imaging across medical institutions, and the expanding knowledge and experience with cardiac computed tomography imaging lead to a growing volume of CT procedures prescribed for cardiac diagnostic purposes. The following review will summarize current applications of cardiac computed tomography.
Depending on the type of CT hardware and image acquisition protocol, contemporary CT systems can deliver a spatial resolution of approximately 0.5 mm and a temporal resolution of 75 to 150 ms. CT imaging of the heart can be performed with and without intravenous injection of contrast agent. If contrast is injected, the typical volume is approximately 50 to 100 ml, delivered at a rate of 4 to 6 ml/s. There are two basic modes of image acquisition. It can either be performed with “prospective triggering”—meaning that the CT system uses the patient’s ECG to trigger release of photons and image acquisition in a predefined segment of the cardiac cycle. Alternatively, “retrospective gating” can be used: X-ray data are acquired continuously over several cardiac cycles, and this allows reconstruction of images for cardiac visualization at any desired time instant during systole or diastole. The combination of data sets which show the heart at various phases of the cardiac cycle, typically 10 to 20, is the basis for “functional” imaging by computed tomography, for example, for the examination of left and right ventricular function or some aspects of valvular disease. A further advantage of “retrospective gating” versus “prospective triggering” is the higher robustness and often somewhat better image quality, while a clear downside is higher radiation exposure.
In order to achieve high image quality, a number of conditions should be fulfilled. A regular heart rate is usually necessary, patients with arrhythmias such as atrial fibrillation are difficult to image, and image quality is inconstant. Since temporal resolution of cardiac CT is limited, a low heart rate with long duration of diastole is advantageous and especially for coronary artery visualization, a heart rate below 65 beats/min, optimally even below 60 beats/min is recommended . This can usually be achieved by oral or intravenous administration of beta blocking agents . A limited number of patients may have to be excluded from cardiac CT examinations because of the inability to follow breathhold commands and to hold their breath for approximately 8 s, or because of pronounced obesity (high body mass increases image noise and thus degrades image quality).
The radiation exposure of cardiac computed tomography examinations varies over a very wide range. If no measures are taken to limit exposure, effective doses of as much as 30 mSv are possible . Based on these values, some authors have extrapolated “risks of malignancy” as high as 0.4% (one induced cancer per 270 cardiac CT examinations for young women ), even though such estimates are highly unreliable and impossible to prove with epidemiologic studies, given the high spontaneous incidence of malignancies in the population (the lifetime risk of cancer is approximately 50% for men and 30% for women) [5, 6]. All the same, it goes without saying that radiation exposure should be kept as low as in any way achievable. During the past years, tremendous progress has been made concerning the reduction of radiation exposure for cardiac CT imaging. Non-contrast-enhanced and relatively low-resolution data sets to detect and quantify coronary calcification can be acquired with effective doses of 0.5 mSv or less. For coronary CT angiography, which requires higher spatial resolution and therefore is associated with higher exposure values, several specific strategies to limit exposure have been developed. They include the use of 100-kV tube voltage instead of the traditional 120-kV tube voltage in patients who are not overweight, the use of tube current modulation to reduce exposure during systole, and the use of prospective ECG triggering instead of retrospective gating. Using modern hardware, the investigation can be performed in many patients with dose values between 2 and 5 mSv in selected cases even below 1.0 mSv [7–11] (for comparison, the effective dose of invasive coronary angiography is approximately 7 mSv ). However, it needs to be clearly stated that doses in clinical practice may be higher, that the use of low- and very low-dose protocols requires expertise as well as up-to-date scanner technology, and that careful patient selection is a further prerequisite for many low-dose image acquisition algorithms.
Morphology and function
After contrast administration, computed tomography permits high-resolution imaging of cardiac morphology. Due to the close to isotropic nature of cardiac CT data sets, images can be generated in any desired plane (see Fig. 1). Clinically, the need for computed tomography as a method for morphologic cardiac imaging is relatively limited, since echocardiography and magnetic resonance imaging will typically provide all necessary information. For specific questions concerning congenital heart disease, in cases of reduced echogenicity in combination with metallic implants that preclude magnetic resonance imaging, computed tomography can serve as a backup imaging modality. For high-resolution imaging in the immediate vicinity of metallic implants (such as valvular prostheses), computed tomography can be superior to echocardiography and magnetic resonance imaging in some cases (see Fig. 2). For the detection and analysis of cardiac calcification (e.g., pericardial calcification), computed tomography is the method of choice (see Fig. 3).
If image data sets are reconstructed at several time points throughout the cardiac cycle, left and right ventricular function can be analyzed with a high degree of accuracy, as numerous comparisons to echocardiography and magnetic resonance imaging have shown [13–16]. Clinically, however, the utilization of computed tomography for this purpose will be an exception (for example, in patients with implanted left ventricular assist devices who may have extremely limited echocardiographic windows and in whom magnetic resonance imaging is contraindicated, see Fig. 4). Cardiac valves, especially the aortic valve, can be visualized with high image quality in computed tomography, and it is possible to quantify the aortic valve orifice area [17, 18]. Again, CT would only be used if all other methods of clarifying the degree of stenosis failed. Imaging the mitral valve (because of its complex anatomy) and especially of the right-sided cardiac valves (because of difficult contrast enhancement) is substantially more difficult by CT than imaging of the aortic valve and surrounding structures.
Coronary artery calcification can easily be detected and quantified by computed tomography. The injection of contrast agent is not necessary, and radiation exposure is very low (see Fig. 5). Coronary calcium is always caused by coronary atherosclerosis (with the only exception of patients on dialysis, in whom coronary arteries can calcify independently from atherosclerosis). The extent of coronary calcium is a surrogate marker for the extent of atherosclerotic plaque in the coronary arteries, and it is easily comprehensible that, consequently, coronary calcium represents a very good prognostic marker concerning future cardiac events such as myocardial infarction and cardiac death. This has been demonstrated numerous times in prospective trials [19–24]. The prognostic power is consistently found to be higher than that of traditional risk factors [21, 25] and higher than that of other surrogate parameters for coronary atherosclerosis, such carotid intima-media thickness . There are no strict guidelines as to which patients should be subjected to a coronary artery calcium scan. Generally, however, it is assumed that coronary calcium detection can be useful in individuals who are asymptomatic and, based on traditional risk factors, at intermediate risk for future cardiovascular disease events, especially if the decision to initiate risk modification measures, such as lipid-lowering therapy, is unclear and requires additional information. Coronary calcium can provide prognostic information and guide therapeutic decisions, even though currently no prospective intervention trials are available. It is important to note that in symptomatic individuals, the presence of coronary artery stenoses in the absence of calcium is unlikely, but not impossible. Hence, the absence of detectable calcium in symptomatic individuals cannot be used to rule out coronary artery stenoses.
Non-invasive coronary CT angiography
After intravenous injection of contrast agent, computed tomography permits visualization of the coronary artery lumen and, hence, “non-invasive coronary angiography” (see Figs. 6 and 7). Currently, 64-slice systems are considered the minimum technical prerequisite for this type of investigation. Subsequent scanner generations provide more robust imaging and a slightly better image quality. As a general rule, careful patient selection and adequate patient preparation is required prior to the examination [1, 27]. If appropriately performed, coronary CT angiography provides reliable coronary visualization in the majority of patients as well as a high degree of reliability for the detection and especially to rule out coronary artery stenoses. In two recent multicenter trials [28, 29], coronary CT angiography was reported to have a sensitivity of 95–99%, specificity of 64–83%, and negative predictive value of 97–99% to identify patients with at least one coronary artery stenosis among individuals at low to intermediate risk for coronary artery disease (see Table 1). Typically, the positive predictive value is lower (64% and 86% in the trials cited above), which is due to a tendency to overestimate stenosis degree in coronary CTA as well as the fact that image artifacts often result in false-positive interpretations. As in any diagnostic test, there is a tradeoff between sensitivity and specificity: Most studies aim to keep sensitivity high, but if a high sensitivity is desired, specificity will suffer—and vice versa. In a multicenter study of 291 patients with 56% prevalence of coronary artery stenoses, as well as 20% of patients with previous myocardial infarction and 10% with prior revascularization, specificity was high (90%) and the positive predictive value was 91% . However, this came at the cost of decreased sensitivity (85%) and negative predictive value (83%, see Table 1). In clinical practice, the more likely approach is to try and keep sensitivity as high as possible, with consequent reduction of sensitivity and positive predictive value.
It is of particular importance to note that ruling out obstructive coronary artery stenoses by coronary CT angiography has prognostic value: Several cohort studies have demonstrated extremely low rates of clinical events after coronary CT angiography had ruled out coronary artery stenoses in patients with stable angina pectoris or acute chest pain [23, 24, 30–34]
Hence, coronary CT angiography can be considered a clinically useful technique in order to rule out the presence of coronary artery stenoses in appropriately selected patients with suspected coronary artery disease, with the goal to avoid further non-invasive or invasive diagnostic procedures. Even though—similar to invasive coronary angiography—a negative test may not provide a conclusive explanation of the origin of the patient’s symptoms, the prognosis is very good if CT fails to demonstrate significant luminal narrowing.
Recently published “Appropriateness Criteria” for the application of cardiac computed tomography lists a number of clinical situations in which coronary CT angiography is considered useful (see Table 2) . Some major areas are the workup of patients with unclear stress test results, or the workup of patients who present to the emergency room with acute chest pain, but a low likelihood of coronary artery disease. A typical goal will be to rule out coronary artery stenoses and in this way avoid invasive coronary angiography. It should be pointed out that excellent image quality is very important to achieve this goal and that decreased image quality typically results in false-positive results, which may lead to unnecessary downstream testing. Hence, selection of appropriate patients, careful preparation, and expert data acquisition are important to maximize the clinical benefits of coronary CT angiography .
The visualization of coronary artery stents is substantially more difficult than assessment of the native coronary artery vessels. Especially for stents of smaller diameter, artifacts are frequent and typically case false-positive interpretations . Hence, coronary CT angiography is usually not recommended in patients after coronary stent implantation (with the only possible exception of large stents placed in the left main coronary artery) .
Bypass grafts can easily and reliably be visualized by cardiac computed tomography. They typically have a larger diameter and move less than the native coronary arteries and are therefore not as challenging to visualize. However, the identification of stenoses at the coronary anastomotic site can be difficult, and a major problem is often severe calcifications, paired with small vessel diameters, of the native vessels in patients with bypass grafts. This decreases accuracy for stenosis detection in the native arteries of patients after bypass surgery (especially the positive predictive value can be rather low) and limits the clinical utility. The current “Appropriateness Criteria”  list the establishment of bypass patency in symptomatic post-CABG patients as a potential application for cardiac computed tomography imaging, but the limitations concerning analysis of native arteries—which will undoubtedly decrease as technology progresses further—need to be considered.
One major area of interest is the visualization of coronary atherosclerotic plaque by computed tomography. As opposed to invasive coronary angiography, coronary CT angiography, a cross-sectional technique, is not restricted to visualization of the contrast-enhanced lumen but also permits visualization of the coronary arterial wall and, hence, of coronary atherosclerotic plaque. Not only stenotic lesions but also non-obstructive plaques, whether calcified or non-calcified, can be visualized if image quality is sufficiently high (see Fig. 8). Initial investigations have been able to demonstrate that non-stenotic plaque seen in coronary CT angiography has prognostic implications [24, 38, 39]: The risk for future cardiovascular events and mortality is increased if computed tomography demonstrates coronary atherosclerotic plaque. Increasing plaque burden is associated with a greater risk . Most data have been obtained in symptomatic patient populations—there is very little information on the prognostic meaning of coronary CT angiography in asymptomatic individuals [40, 41]. It has also been demonstrated nicely that, again, the negative predictive value of CT is particularly high: Event rates are superbly low when no plaque is present. The positive predictive value is not as high: Even in the presence of plaque, event rates are relatively low , which may have to do with the fact that coronary CT angiography is typically applied in low-risk populations.
Several authors have been able to show that the ability of CT extends beyond mere plaque detection and quantification: Certain plaque features, such as positive remodeling (see Fig. 9), spotty calcification, and low attenuation values (typically, below 30 HU), are associated with a particularly high risk of experiencing a future cardiovascular event .
Support of interventional procedures
Cardiac computed tomography, through its ability to provide high-resolution, three-dimensional, and isotropic morphology, is increasingly being used as a method to support interventional procedures. For the interventional treatment of chronic total coronary artery occlusions, information provided by CT angiography on the course and length of the occluded vessel segment, as well as the extent of calcification, has been found help with the recanalization procedure and to provide information on the likelihood of success [44–46]. In patients who are evaluated concerning transcatheter aortic valve implantation, CT can provide valuable information for the exact analysis of dimensions and spatial configuration of the aortic root complex (see Fig. 10) [47–50]. In this context, the ability of CT imaging to predict suitable fluoroscopic angulations for monitoring of the implantation procedure is of importance. If suitable fluoroscopic angulations are obtained from a pre-procedure CT, contrast exposure due to repeated aortograms which serve only to identify a good angulation of the C-arm can be substantially reduced [51, 52]. Furthermore, computed tomography has the ability to serve as anatomic reference for electrophysiologic ablation procedures  and is often used for this purpose in clinical practice.
Multi-center studies that addressed the accuracy of coronary artery stenosis detection by contrast-enhanced 64-slice coronary CT angiography
Number of sites
Number of patients
Prevalence of obstructive CADa (%)
Negative predictive value (%)
Positive predictive value (%)
List of appropriate indications for cardiac computed tomography, according to an Expert Consensus Document published in 2010 (modified according to )
Coronary CT Angiography
Stable, Suspected CAD
Testing for ischemia not possible, equivocal, or discordant to symptoms
Stable, Suspected CAD
Intermediate pre test likelihood of CAD (10–90%)
Acute Chest Pain
ECG and enzymes normal or equivocal
Low to intermediate likelihood for CAD
Planned Non-coronary Cardiac Surgery
Intermediate likelihood for CAD
Progressive Chest Pain
Earlier testing for ischemia was negative
New Heart Failure
Low to intermediate likelihood for CAD
Bypass Patency if post-CABG with Chest Pain
Localization of Bypass Grafts and other Retrosternal Anatomy
Prior to reoperative cardiac surgery
Left Main Stent ≥3.0 mm, Asymptomatic
Evaluation of Coronary Anomalies
Noncontrast Coronary Calcium Imaging
Risk Assessment in Asymptomatic Low Risk Individuals
With a positive family history of premature coronary artery disease
Risk Assessment in Asymptomatic Individuals at Intermediate Risk
Evaluation of Cardiac Structure and Function
Pulmonary Vein Anatomy
Prior to radiofrequency ablation
Coronary Vein Mapping
Prior to biventricular lead placement
Assessment of Complex Congenital Heart Disease
Assessment of Right Ventricular Dysplasia
Inadequate Images from Other Modalities
Cardiac valves in suspected dysfunction
Valvular prostheses in suspected dysfunction
Suspected cardiac masses
Assessment of left and right ventricular function
In summary, cardiac computed tomography has a growing number of clinical applications, the most important of which is non-invasive coronary angiography. Coronary CT angiography should only be used when patient characteristics promise good and fully diagnostic image quality and in settings where pre-test likelihood is not very high (otherwise, invasive angiography with the option for immediate revascularization would be more appropriate). In other areas, such as visualization of cardiac morphology and especially function, cardiac CT is less frequently useful and appropriate only in selected cases. It can be expected that the technical development of computed tomography will continue at a rapid pace and that, consequently, clinical applications of cardiac computed tomography will maintain further growth.