Introduction
Noninvasive assessment of the coronary arteries using multidetector CT (MDCT) represents one of the most promising diagnostic imaging applications in contemporary cardiology. Single-center and multicenter studies have confirmed the accuracy of 64-slice MDCT in detection of obstructive stenosis, relative to the clinical gold standard of invasive coronary angiography.1, 2, 3, 4 This application has fueled tremendous technology advancement in hardware and software systems over the past decade that have led to substantial improvements in image acquisition, reconstruction, work flow and analysis. These improvements have not only advanced the status of MDCT coronary angiography, but have also enabled a number of complementary noncoronary myocardial applications that can potentially be performed at the time of coronary evaluation in a single exam. However, radiation exposure is a primary concern with the increasing use of MDCT technology in diagnostic cardiac imaging and has become a centerpiece of new hardware, software and imaging protocol improvements. Additionally, MDCT acquisition methods that use partial coverage detectors are limited by long scan times, image artifacts, and high radiation and dose of iodinated contrast agent. The advent of wide detector designs that offer full cardiac coverage provide advantages that can potentially improve these shortcomings and facilitate advanced myocardial imaging applications. Although these applications are at early stages of development and much of the published data has been obtained in preclinical models, early clinical experience has produced encouraging results.
In this Review, we provide an update on technological advances in MDCT imaging developed in the past 5 years, with a focus on newly enabled complementary myocardial applications. Specifically, we present recent data demonstrating the feasibility of MDCT delayed contrast-enhanced myocardial viability assessment, as well as combined MDCT coronary angiography and myocardial perfusion imaging. Lastly, we discuss new CT contrast agents to facilitate the development of CT molecular imaging. Given that the use of MDCT for the assessment of cardiac volumes, global and regional ventricular function has been reviewed extensively elsewhere,5, 6, 7 this aspect is not covered in this Review.
Advances in CT scanner technology
Early gantry-based CT scanners in the 1970s and 1980s used modified X-ray tubes and detector movement that resulted in a system with a fixed detector array that could obtain images with a scan time of 2 s.8 Although imaging of cardiac chambers was possible,9 the temporal and spatial resolution of these systems limited their applicability to coronary imaging. A major breakthrough in CT technology came with the development of slip-ring technology,10 which allowed the gantry mechanism to continuously rotate around the patient while the table was advanced. This technology led to the development of helical CT in the late 1980s and introduced a new approach to obtain three-dimensional volume data. The widespread introduction of single-slice helical CT systems (first generation scanners) in 1996 marked a turning point for cardiac CT applications. Advanced reconstruction methods required only 180° rotation of the gantry, thereby improving temporal resolution to 400 ms. Helical-based data acquisition led to a renaissance of CT and to renewed investments in CT technology, particularly to the development of multiple detector elements, which provides the basis for faster scan times, wider volume coverage and high spatial resolution demonstrated by modern systems (Table 1, Figure 1).11 Third generation single-source 64-slice MDCT scanners deliver isotropic resolution as low as 0.35 mm with a temporal resolution of 165 ms (for half scan reconstruction) and have been tested extensively for accuracy, relative to conventional coronary imaging.1, 3 Although these studies showed a glimpse of the ultimate promise for CT coronary imaging, they also revealed a number of sobering limitations related to spatial and temporal resolution, long scan times, limited ability to image patients with arrhythmia, image artefacts and high radiation dose. These limitations are especially apparent in cases evaluating patients with in-stent stenosis, severely calcified plaque and multicomponent plaques.1, 2, 3, 4 Attempts to solve these problems led to fourth generation scanners in 2007, where, for the first time, detector designs and software advances implemented by CT manufacturers diverged into three distinctly different improvement targets: temporal resolution, detector coverage and spatial resolution with spectral imaging.
Figure 1 | Technical progression of scanner technology.
Improvement in coverage of the z axis with 4-slice, 16-slice and 64-slice detector rows, relative to wide-range 320-slice detector rows. With the wide-range technology, the entire heart is covered in one gantry rotation.
Dual-source MDCT: temporal resolution
Temporal resolution remains one of the most important limitations for coronary CT imaging. Temporal resolution of 100–200 ms can produce diagnostic exams, whereas approximately 50 ms is needed to completely freeze cardiac motion and eliminate the need for heart rate control.11 However, at a gantry rotation speed of 300–350 ms, MDCT scanners are approaching current engineering limits for gravitational forces on the gantry. An alternative approach to increase temporal resolution is dual-source MDCT technology, instead of further increasing the gantry rotation speed. Although this concept was first proposed in the late 1970s,12, 13 almost 30 years elapsed before its implementation into clinical practice.
In dual-source MDCT, two acquisition systems are mounted onto the rotating gantry with an angular offset of 90°. Both detectors allow for the simultaneous acquisition of 64 overlapping, 0.6 mm slices by means of double z sampling, which makes use of a periodic motion of the focal spot in the longitudinal direction to improve data sampling along the z axis. By continuous electromagnetic deflection of the electron beam, the focal spot is wobbled between two different positions, and the amplitude of the periodic z motion is adjusted in such a way that two subsequent readings are obtained on 32-slice detector system, which are then combined to a 64-slice projection.14
Initial studies show that dual-source MDCT in conjunction with half-scan reconstruction yield a true temporal resolution of 83 ms, which enables cardiac imaging at heart rates above 60 bpm and reduces the occurrence of artifacts caused by coronary calcification.15, 16, 17 MDCT images can be acquired at elevated and variable heart rates without a negative effect on the overall image quality.18, 19 The use of two X-ray sources initially raised concerns about radiation exposure; however, reports have indicated that the effective radiation dose can be reduced to 1.2–9 mSv, relative to single-source MDCT.20, 21
Dynamic-volume CT: detector coverage
With the introduction of wide-range detector CT systems, the traditional principles of helical scanning by combining volume data from subsequent spiral acquisitions to cover the whole heart are abandoned. Wide-range MDCT imaging allows the entire heart to be acquired in a single gantry rotation (Figure 1), without the need for table movement, which enables dynamic volume imaging. Wide-range detectors overcome the problem of slab misregistration artefacts that occur when adjacent volumes are acquired at two different heartbeats. As the entire volume of the heart is acquired in a single heartbeat, beat-to-beat variations in heart rate do not present a problem. The single gantry rotation ensures temporal uniformity of the acquired volume data and thereby equal contrast distribution (contrast uniformity) throughout the cardiac data set, which facilitates interpretation of diagnostic findings.22, 23
Dynamic volume acquisition with wide-range detector CT system can reduce the radiation exposure by 4–5-fold, compared with traditional helical imaging, because overscanning and overranging are not required.23 Wide-range MDCT cover an area of up to 160 mm in the z direction and, therefore, greater cone beam geometry has to be reconciled. If the beam shape is not properly accounted for in the reconstruction, the image quality can become compromised. However, initial clinical data show high diagnostic accuracy for detection of coronary artery stenosis with wide-range detector CT system compared with conventional coronary angiography.23 Moreover, dynamic volume acquisition with wide-range MDCT has clinical potential to develop four-dimensional acquisition strategies for myocardial blood flow assessment complementary to coronary angiography.22
Dual-energy and spectral CT
X-ray tubes in the current generation of MDCT scanners produce photons with a range of energies so called polychromatic X-ray beams. Photons of different energies possess differing spectra and, therefore, differing absorption profiles.24 Thus, imaging the same volume of multicomponent tissue at different energies can provide detailed tissue characterization and differentiation by CT. Separation of material such as calcium, iodine and water is facilitated and beam hardening artifacts are reduced. The potential to use spectral information from X-ray beams was first reported in the mid 1970s.25 Now, dual-energy applications using either dual-source MDCT or single-tube MDCT systems with ultrafast dual-energy switching technology in combination with specially designed high-definition high-resolution detectors are commercially available. Energy discrimination on the detector level (spectral MDCT) is under intensive investigation and is likely to be presented soon for clinical use. The greatest benefit of spectral decomposition in cardiovascular applications is that it permits the detection and potentially quantification of several distinct signals, which may have value in plaque characterization and perfusion imaging (Figure 2).24, 26, 27, 28, 29, 30 These potential benefits have sparked renewed interest in this technology. Table 2 summarizes the current technical approaches that take advantage of dual-energy and spectral CT techniques.
Figure 2 | Dual energy acquisition of perfusion deficits (black arrows) are shown as short-axis multiplanar reformation.
Corresponding 1.5 mm sections acquired simultaneously with the a | low-voltage and b | high-voltage tube, reconstructed with 165 ms temporal resolution. c | Calculated iodine map and d | merged image using 5 mm reconstructed section thickness. e | 50% overlay and f | 70% overlay of the iodine map and the merged reconstruction show the area of septal hypoperfusion. Permission obtained from Elsevier Ltd © Schwarz, F. et al. Eur. J. Radiol. 68, 423–433 (2008).
Advances in imaging protocols
Radiation dose reduction remains one of the most important goals of cardiac MDCT imaging. Radiation exposure is determined by a number of interdependent technical factors, including the beam energy settings, collimation, pitch and duration of beam exposure. These parameters also determine the quality of data contained in the target volume, particularly the ability to resolve differences in contrast.31 Interest has increased in protocol optimization as a method to minimize the duration of beam exposure and thus total radiation dose. Tomographic imaging modalities offer the possibility of retrospective and prospective electrocardiogram (ECG)-triggered image acquisition. In retrospective imaging, the whole R-to-R' interval is acquired, whereas prospective image acquisitions are restricted to predefined period of the cardiac cycle, usually the end-diastolic phase.
Retrospective ECG-triggered imaging
With acquisition of whole R-to-R' interval, images can be reconstructed from any phase of the cardiac cycle, and the best cardiac phase can be chosen for diagnostic evaluation (Figure 3a). Until 64-slice technology was introduced, volumetric coronary artery imaging was performed with retrospective ECG-triggered imaging protocols. Retrospective imaging also offers the opportunity for global and regional functional analysis at high quality.11, 32 To improve temporal resolution, multisegment reconstruction methods that use partial data acquisitions from several consecutive heartbeats to generate complete volume data sets have been applied.32, 33 However, this technology offers only limited improvement in image quality at heart rates above 65 bpm, while adding additional radiation to the imaging protocol (up to 25 mSv1, 3, 34, 35).
Figure 3 | Examples of CT coronary angiography ECG-gating protocols for the 320-slice detector system.
MDCT protocols designed for coronary artery and functional imaging require X-ray exposure and imaging data acquisition over one heart beat with the R–R' interval flanked by additional ms before and after the R–R' interval. With retrospective imaging protocols a | the tube current can remains uniform over the imaging interval, or, b | dose modulation can be used to reduce tube current during systole and increased during diastole aimed at reducing the overall radiation dose. c | Prospective ECG-gating enables coronary artery imaging with the X-ray exposure only in a 10% portion of diastole or less over one heart beat. d | Alternatively, to optimize temporal resolution, multisegment acquisition and reconstruction can be performed in a portion of diastole over several heart beats in 320-slice wide-range detector CT systems. Abbreviations: ECG, electrocardiogram; MDCT, multidetector CT.
The introduction of tube modulation techniques for retrospective ECG-triggered imaging reduced radiation dose by 20–30%.35 In this approach, systolic phases of the cardiac cycle are acquired with a low tube current that is then rapidly increased in the diastolic phase for coronary imaging (Figure 3b). This method allows acquisition of both high quality coronary imaging and myocardial motion assessment with a reduced, but acceptable, image quality. In addition, imaging protocols can be tailored to patient-specific parameters with the use tube current modulation. The tube current is adjusted to the patient's body build and weight, which can reduce radiation dose to 10–15 mSv in cardiac exams.35
Prospective ECG-triggered imaging
If functional myocardial information is not required, modern MDCT systems permit prospective ECG-triggered imaging for radiation dose reduction. In prospective ECG-triggering, the image acquisition is limited to a narrow, predefined end-diastolic phase, which results in a reduction of radiation exposure to 1–3 mSv (Figure 3c).36, 37 During CT coronary angiography imaging with prospective ECG-gating, data are acquired with z coverage of only 40 mm in 64-slice systems. Each location that is covered by the detector range represents a volume of the patient's anatomy. In these prospective ECG-triggered, so-called 'step and shoot' acquisitions, each volume is taken while the table is stationary; the table is then moved from one location to the next, until the whole heart volume is obtained. Initially, datasets were acquired at every second heartbeat to allow time for table movement and for the contrast bolus to travel, but this strategy results in differing contrast medium concentrations at each imaging location.36 Importantly, low heart rates of about 60 bpm are required to ensure sufficient image quality for diagnostic assessment. High-pitch acquisitions in dual-source systems using the prospective ECG-triggered technology with fast table movement and high temporal resolution have been reported suggesting the feasibility of coronary CT angiography acquisitions with dose values below 1 mSv.38, 39
In contrast to 64-slice systems, dynamic volume acquisition with 320-row wide-range MDCT can use multisegment acquisition and reconstruction in the prospective ECG-gated mode (Figure 3d). Using prospective ECG-gated cardiac CT protocols limited to 10% of the R–R' interval on a 320-row detector system, the radiation doses range is 4–5 mSv, but acquisitions as low as 1.7 mSv can be achieved.40 For coronary assessment, single-beat, two-beat and three-beat acquisitions are recommended when heart rate is less than 65 bpm, 65–75 bpm and greater than 75 bpm, respectively. However, with each additional acquired beat, temporal resolution is improved at the expense of increased radiation exposure.23, 40
New applications
The technical advances described have been driven mainly by the need for improved imaging quality for coronary vessel assessment. Until recently, MDCT imaging offered information restricted to anatomy and ventricular function. The technical development has also sparked the interest in complementary myocardial applications to broaden the use of cardiac CT and improve diagnostic accuracy for coronary artery disease. These new applications characterize and measure biological processes on a cellular level, providing information on physiological mechanisms in the myocardium. The rapid progress has now placed these applications within reach for clinical use.
Myocardial viability
The potential role of CT for the assessment of myocardial viability and the detection of myocardial infarction was first described by Gray et al. in studies of explanted dog hearts in the late 1970s.41, 42 In a series of elegant ex vivo studies, Higgins et al. were able to show the preferential uptake of contrast media in myocardial areas damaged by acute ischemia.43, 44 Data accumulated, demonstrating that acute infarcts are detectable by CT.45, 46 Although these initial results were encouraging, the practical use of CT technology was hampered by insufficient in vivo image quality. Modern MDCT systems are well suited to evaluate myocardial viability in combination with intravenous contrast agents. However, a number of challenges remain, relating to the required contrast dose, image quality and radiation exposure.
First-pass and delayed-enhancement imaging
In general, two approaches for visualizing damaged myocardial tissue by MDCT have been reported: first-pass acquisitions immediately after intravenous delivery of the contrast agent and delayed contrast-enhancement imaging several minutes after contrast injection. First-pass imaging has been used to detect and quantify myocardial infarcts;47, 48, 49, 50 however, these acquisitions reflect myocardial blood flow and are therefore, strictly speaking, detection of perfusion deficits not tissue viability. First-pass imaging relies on the administration of iodinated contrast agents, which remain primarily in the intravascular compartment during the early phase after intravenous delivery.51, 52 The reduced or absent myocardial blood flow can be appreciated as a hypoenhanced area compared with normal myocardium during the first-pass of the iodinated contrast agent (Figure 4). Initial studies in patients using 16-slice MDCT technology showed that, in the setting of acute reperfused infarcts, perfusion deficits tends to underestimate infarct size in MDCT, whereas delayed iodine-enhancement MDCT images showed good agreement with MRI.49 These results have been reproduced with 64-slice MDCT technology.50
Figure 4 | Time course of contrast enhancement in an acute myocardial infarction after intravenous iodine injection.
An axial temporal image series demonstrates postreperfusion contrast agent kinetics after 150 ml injection of contrast agent (panels a–h). a | First-pass image during contrast-agent injection. The signal density of the infarct in the first pass is substantially lower than that of the remote myocardium. b | 5 min after injection, the signal density of the damaged myocardial region is markedly greater than that of the remote myocardium. c | The damaged myocardium still appears bright 10 min after contrast delivery, while signal intensity in the remote myocardium and in LV chamber is lower. d–h | The washout of contrast can be visually appreciated 20 min after contrast injection and is most apparent in the left ventricular chamber, which represents the distribution of iodine contrast in the blood pool and in the remote myocardium. i | Quantitative contrast kinetics for the left ventricle, remote myocardium, and infarct after injection of contrast agent. As can be seen in b, the infarct becomes well delineated and reaches peak enhancement at 5 min after injection and then washes out in proportion to the chamber (blood pool) and remote myocardial signal. Permission obtained from Lardo, A. C. et al. Circulation 113, 394–404 (2006).
Delayed contrast-enhanced MDCT imaging is based on the principals of delayed enhanced MRI to image the myocardium several minute after contrast delivery to evaluate tissue viability.64 Studies in preclinical animal models have demonstrated a good correlation between delayed contrast-enhanced MDCT imaging and infarct size and morphology, including assessment of microvascular obstruction patterns, compared with postmortem pathology.50, 53, 54, 55 In addition, evaluation of patients with previous myocardial infarction confirmed preclinical data that the anatomical extent of myocardial infarction assessed by MDCT reflects nonviable tissue.50, 56 An excellent correlation of delayed contrast-enhanced MRI and delayed contrast-enhanced MDCT has been established for acute,50, 54, 56 and chronic,55, 56, 57 infarcts (Figure 5, Table 3). In addition, with its high isotropic resolution, MDCT has the potential to greatly decrease partial volume effects and accurately characterize tissue compositions and dynamic remodeling processes as it has been demonstrated recently for the peri-infarct zone.55
Figure 5 | Examples of coronary-MDCT and DE-MDCT in three representative patients.
The MDCT images are associated with corresponding coronary angiography and DE-CMR images. a | No coronary artery disease and absence of delayed contrast enhancement. b | No coronary artery disease and midventricular delayed contrast enhancement. c | Occlusion of the proximal LAD with transmural delayed contrast enhancement in the apex, septum and anteroseptal region and mural thrombus. The proximal LAD, anteroseptal region and mural thrombus are indicated by yellow, red and green arrows, respectively. Abbreviations CMR, cardiovascular magnetic resonance; DE, delayed enhancement; MDCT, multidetector CT; LAD, left anterior descending artery. Permission obtained from the European Society of Cardiology © le Polain de Waroux, J.-B. et al. Eur. Heart J. 29, 2544–2551 (2008).
Delayed contrast-enhanced MDCT has also been applied in hybrid imaging approaches with nuclear techniques,58, 59 and in a multimodality approach with invasive coronary angiography. Habis et al., for example, took advantage of the readily delivered iodine during invasive coronary angiography in patients with acute coronary syndrome, and demonstrated that 64-slice delayed enhanced MDCT imaging without iodine reinjection immediately after an invasive catheter procedure is a promising method for early viability assessment.60 Sato et al. used the same multimodality approach to establish the prognostic value of delayed contrast-enhanced MDCT.61 Imaging protocols and technical details of human and animals studies exploring MDCT delayed enhanced viability imaging after myocardial infarction are summarized in Table 3. The capability of MDCT to detect myocardial delayed enhancement in patients suffering from acute myocarditis has been reported as well.62 Thus, delayed enhancement with iodine contrast seems to be specific for irreversible cell damage, irrespective of etiology. Although studies have been conducted in patients (Table 3), a reduction in radiation dose for a complementary scan is essential for routine clinical application. Results from our laboratory show that prospective ECG-gated delayed contrast-enhanced MDCT with low radiation exposure provides high-resolution imaging of acute myocardial infarction while lowering the radiation dose to 0.72 mSv.63 Prospective ECG-gated MDCT imaging allowed an accurate assessment of infarct size and microvascular obstruction pattern, when compared with retrospective ECG-gated MDCT imaging (Figure 6).
Figure 6 | Comparison of prospective and retrospective ECG-gated MDCT images obtained 10 min after injection of contrast agent.
Visualization of myocardial infarction (white arrow) 10 days after myocardial infarction in retrospective and prospective scans, with (black arrow) or without MVO. Abbreviations: ECG, electrocardiogram; MDCT, multidetector CT; MVO, microvascular obstruction. Permission obtained from Elsevier Ltd © Chang, H. J. et al. J. Am. Coll. Cardiol. Imaging 2, 412–420 (2009).
Mechanisms of delayed contrast enhancement
Minutes after intravenous administration of an iodinated contrast agent, delayed hyperenhancement of infarcts becomes apparent (Figures 4, 5, 6).53 Delayed contrast-enhancement imaging is feasible with MDCT because iodinated contrast agents are biologically inert and diffuse passively into the increased extracellular matrix of infarcted myocardium with a distribution half-life of approximately 20 min. In general, the magnitude of myocardial enhancement is dependent on local myocardial blood flow and vascular integrity, local distribution volume of the contrast agent, and intrinsic properties of extracellular contrast agent. In the setting of acute myocardial infarction, delayed enhancement is explained by myocyte necrosis, sarcomere membrane rupture, and passive diffusion of contrast into the intracellular space.64 The mechanism of delayed hyperenhancement in chronic collagenous myocardial scar tissue is thought to be related purely to an increase in contrast medium in the interstitial spaces between collagen fibers and, thus, a larger volume of contrast distribution in the scar tissue, compared with that of tightly packed myocytes. Distribution of contrast material in the collagenous scar is highly dependent on the wash-in and washout kinetics of the contrast agent. The timing, therefore, between administration of contrast agents and imaging is crucial to accurately assess chronic infarct size in MDCT imaging. With respect to detailed tissue characterization for myocardial viability imaging, an important feature of MDCT is that signal density values are unique and determined by the physical properties of individual constituents of the heart, including blood, viable and nonviable myocardium, that result from direct attenuation of the X-ray beam by iodine molecules. This is in contrast to enhancement mechanisms by MRI that rely on alteration of the contrast media via interactions with water molecules.64
Myocardial perfusion
Measurement of myocardial perfusion by CT was first attempted in the late 1980s with the invention of dynamic electron beam tomography.65, 66 In contrast to electron beam tomography, helical MDCT cannot fully measure all phases of myocardial contrast kinetics because volume acquisition is not dynamic in helical imaging protocols (Figure 7). However, by capturing myocardial enhancement during a specific phase of the first-pass contrast curve, George et al. were able to show differences in myocardial perfusion with iodine contrast during helical MDCT imaging (Figure 7a,b). These experiments provided the evidence that helical MDCT using retrospective ECG-triggered coronary angiographic imaging protocols is capable of detecting myocardial perfusion patterns, and that myocardial signal density is an accurate surrogate measurement for myocardial blood flow.67
Figure 7 | Helical and dynamic MDCT perfusion imaging of a canine model of LAD stenosis.
Reconstruction of helical acquisitions in a | short axis and b | long axis showing perfusion deficits during infusion of adenosine. c | Signal density–time curves of intravenous iodinated contrast agent during helical and dynamic MDCT scanning. The gray shaded area illustrates the period during which helical MDCT scanning took place. The entire signal density curve is generated from dynamic MDCT acquisition. Myocardial curves were measured from the anterior myocardial wall (stenosed), the lateral myocardial wall (remote), and the left ventricular blood pool. d | Adenosine stress dynamic MDCT imaging of the mid left ventricle at 0, 15, 21 and 70 s. Note the area of hypoenhancement (arrows) in the anterior myocardial wall as contrast first arrives in the left ventricle (15 s and 21 s). No visually significant differences are noted at the end of imaging (70 s). The continuous signal density measurements in the areas of white circles (anterior and the lateral walls) and red circle (left ventricular blood pool) are graphically shown in panel c. Abbreviations: LAD, left anterior descending artery; MDCT, multidetector CT. Permission obtained from Wolters Kluwer Health © George, R. T. et al. Invest. Radiol. 42, 815–822 (2007).
Absolute myocardial blood flow measurement (ml/min/g) is possible with MDCT technology, but remains an area of investigation. Studies from our laboratory have confirmed that dynamic MDCT can accurately quantify myocardial blood flow using upslope and model-based deconvolution methods when compared with microsphere derived myocardial blood flow values.68 However, the feasibility of dynamic MDCT imaging protocols has not yet been demonstrated and may not be applicable in the clinical setting because of the continuous radiation exposure during the first pass of the contrast agent (Figure 7c,d).
Early clinical experience shows that semiquantitative parameters for myocardial tissue perfusion can be applied to evaluate coronary reserve and compare favorably with single photon emission CT (SPECT) imaging.22 Functional capacity of coronary collaterals, physiological effectiveness of therapeutic interventions, and the physiological importance of obstructive coronary lesions can be analyzed in a similar fashion. Several clinical reports have been published in the past year, which evaluate patients with CT perfusion deficits in myocardial territories with matching coronary pathology and show good agreement with MRI and SPECT.22, 29, 48, 50 Cury et al. showed that patients with acute myocardial infarction can be identified on the basis of regional wall motion abnormalities and MDCT perfusion deficits.50 The clinical imaging protocol applied in this study represented a comprehensive noninvasive MDCT imaging exam, including coronary imaging, function and perfusion imaging.
MDCT myocardial perfusion imaging as an adjunct to CT coronary angiography has been demonstrated in preclinical and clinical studies using 64-slice detector systems.22, 67 Although the results of these studies are promising, combined MDCT angiography and perfusion imaging using the 64-slice detector systems has limitations. Depending on the manufacturer, 64-slice detector systems are limited to cardiac coverage of 40 mm or less in the z axis. Given the radiation and contrast doses required for retrospectively ECG-triggered 64-slice MDCT cardiac imaging, acquisition of resting and stress perfusion is limited by the consideration of these two parameters. The use of wide-range detector MDCT enables resting and stress perfusion imaging with acceptable levels of radiation exposure and contrast dose. Imaging of resting perfusion not only allows the chance to evaluate the reversibility of a perfusion deficit, but also allows a resting CT angiogram to be performed without adenosine and, therefore, avoiding poor image quality attributable to adenosine-mediated tachycardia. After the resting CT angiogram, adenosine stress perfusion imaging using multisegment acquisition and reconstruction can be performed (Figure 8). Full cardiac coverage, dynamic wide range-MDCT systems, allows for simultaneous perfusion imaging of the entire myocardium (temporal and contrast uniformity) with advantages discussed. The feasibility of dynamic wide-range MDCT systems perfusion imaging for the simultaneous evaluation of coronary atherosclerosis and its physiological significance with a mean radiation dose of 13.5 
3.5 mSv (Figure 8) was demonstrated recently.22 However, further methods of radiation dose reduction will be required to make CT perfusion imaging a commonly used clinical application. The international multicenter CorE320 study will compare the combination of CT perfusion and coronary artery assessment with SPECT imaging and conventional coronary angiography. Enrolment for this 2-year study began in September 2009.
Figure 8 | Wide-range detector MDCT perfusion imaging protocol.
a | During the scan preparation, intravenous access is obtained. ECG electrodes are placed and the patient's heart and vitals are monitored. If heart rate reduction is necessary, the patient is prepared for the scan with intravenous hydration and 
-blocker administration. The injection rate for iodine contrast is 5 ml/s. The parameters for stress CTP are as follows 128 
1.0 mm detector collimation, 120 kV, 100 mAs, 3 gantry rotations. CTA and rest CTP are obtained during the same acquisition 10 min after the adenosine was stopped. The scan parameters are as follows 256 0.5 mm detector collimation, 120 kV, 175 mAs, 3 gantry rotations. b–g | Resting and stress wide-range MDCT myocardial perfusion imaging. First-pass, rest and adenosine augmented wide-range MDCT myocardial perfusion imaging in a patient referred for invasive angiography after SPECT showed a partially reversible perfusion deficit in the inferior and inferolateral territories. Wide-range MDCT images are acquired during one heart beat, providing full heart coverage and temporal uniformity of all myocardial segments. Panel b shows a partial reversible perfusion deficit in the inferior and inferolateral wall on radionuclide myocardial perfusion imaging in this patient with exertional angina. The upper panels show stress perfusion imaging and the lower panels show resting perfusion. Rest (panel c) and stress (panel d) MDCT perfusion imaging shows a reversible subendocardial perfusion deficit in the inferior and inferolateral walls (black arrows). Noninvasive angiography (panels d–f) confirms a major stenosis (white arrows) in the proximal right coronary artery (panels e and f) and the proximal left circumflex artery (panel g). Abbreviations: CTA, CT coronary angiography; CTP, perfusion CT; ECG, electrocardiogram; MDCT, multidetector CT; SPECT, single-photon emission CT. Permission obtained from George, R. T. et al. Circ. Cardiovas. Imaging 2, 174–182 (2009).
Dual energy MDCT acquisitions evaluating perfusion deficits have been compared with SPECT (Figure 2).29 The authors of the study suggest that dual-energy-based iodine mapping might be more sensitive for the detection of hypoperfused areas in the myocardium, compared with single-source hypoattenuation signal density values based on the Hounsfield scale. However, the benefits of dual-energy perfusion MDCT, including blood flow assessment, still need to be studied systematically.
Molecular CT imaging
As most disease-specific molecular and cellular processes involve low concentrations in the millimolar to nanomolar range, and show low capacities of accumulating in organ specific sites, images are typically obtained using shortlived isotopes. Molecular and cellular imaging therefore requires probes and imaging methods with sensitivity for detection and quantification of molecular and cellular targets. Although hybrid imaging approaches with SPECT/CT and PET/CT currently show the most-successful results in terms of merging cellular signals with anatomical structures,69 MDCT has the capability to provide both morphologic and molecular information simultaneously in a single scan.
Attempts to visualize cellular processes by X-rays using iodine-based contrast agents have been reported since the late 1970s but have been disappointing.70 At the present time, commercially available iodinated contrast agents that specifically target disease processes are not available because of the difficulty in conjugating iodine to most biological components, a very rapid clearance from the blood, and low contrast sensitivity. A minimum iodine concentration of 0.5 mg/ml is required to achieve a detectable change of about 10–15 Hounsfield units (HU) and ten times this concentration is desirable for molecular imaging.70
However, a recent study showed that an iodine suspension formulation composed of crystalline iodinated particles dispersed with surfactant reached concentrations of up to 67 mg iodine/ml and accumulated in macrophages of atherosclerotic lesions 2 h after intravenous delivery (Supplementary Figure 1).71 Imaging of the atherosclerotic plaques was enhanced more than threefold, compared with conventional iodine contrast agents, and showed a signal more than 10 HU compared with nondiseased vessels. Although these results are promising, the iodine suspension tested follows a passive pattern of macrophage uptake and requires two timely and separate acquisitions, which can cause image misregistration.71
Another approach under investigation is the use of dense nanoparticles containing high atomic number elements as contrast agents for X-ray CT imaging. Gold, for example, offers physical and pharmacokinetic advantages over the currently available iodine-based agents. The K-edge of gold at 80.7 kV, compared with that of iodine at 33.2 kV, confers higher absorptivity.72 Hence, gold provides about 2.7 times greater contrast per unit weight than iodine. Gold nanoparticles have been developed as in vivo X-ray contrast agents and have been applied successfully in small animals.72 Good contrast-to-noise images can be obtained at gold concentrations as low as of 100 
g/ml.73 An antibody targeted molecular imaging approach applying gold nanoprobes for detection of cancer cells with clinical MDCT has been reported.74 The data presented are also relevant to the use of gold nanoparticles as drug-delivery vehicles for chemotherapy, as the biodistribution of gold will determine the drug deployment. However, some caution is required for the human applications using gold nanoparticles as clearance of most nanoparticles is slow. Thus, whole-body retention should be considered, particularly when using nanoparticles for screening purposes.
Gold-containing agents could also be used for direct labeling of cells intended for intramyocardial delivery. Initial experiments show that gold labeling did not affect cell viability over a period of 1 week. Gold-labeled mesenchymal stem cells were readily detected on a 64-slice MDCT clinical scanner at a minimum concentration of ten thousand cells A point-source injection of one million gold-labeled mesenchymal stem cells were visible immediately and 2 weeks after injection in rabbit hindlimbs using both X-ray fluoroscopy and CT.75
Conclusions
Recent advances in MDCT technology have addressed many of the unique and demanding challenges of imaging the heart and have enabled new functional myocardial applications that complement its coronary imaging capabilities with clinically acceptable radiation and contrast doses. However, development of these applications is in the early stages and will require a broad interdisciplinary effort (cardiologists, radiologists, engineers and physicists) to ensure continued advancement. Results from preclinical and single-center clinical studies have shown great promise and these applications are now being rigorously evaluated in large international multicenter clinical trials. While we await the results, designs of the fifth generation scanner continue, which will try to incorporate multiple technologies (for example, dual-source, wide-detector, spectral imaging) into a single imaging system capable of motion-free, high-resolution, spectral evaluation of coronary arteries and physiologic myocardial assessment in a single heartbeat.
Review criteria
Articles cited in this Review were selected from PubMed database searches. Keywords included "computed tomography", "viability", "perfusion", "detector designs". Search words also included known authors. Papers cited include English language articles published in the past 5 years, as well as some historical references. All references are full-length manuscripts.
