Chapter 17 Essentials of Imaging and Imaging Technologies Related to Arrhythmias

A. Cardiac Computed Tomography Jasbir Sra

B. Cardiac Magnetic Resonance Imaging Rishi Anand, Timm-Michael Dickfeld

C. Transthoracic and Transesophageal Echocardiography Anjlee M. Mehta, Navin C. Nanda

A. Cardiac Computed Tomography

Introduction

Fluoroscopy does not provide adequate anatomic visualization because of lack of contrast between the area of interest and surrounding structures, making it difficult to precisely manipulate intracardiac catheters in complex three-dimensional left atrial (LA) anatomy in procedures such as atrial fibrillation (AF) ablation.

Recent advances in imaging technology are starting to have a profound effect on the practice of electrophysiology. Radiological scans, such as computed tomography (CT) and magnetic resonance imaging (MRI), offer high-quality anatomic visualization with high spatial and temporal resolution and can thus enhance efficacy and reduce the risks associated with procedures such as LA ablation for AF through more precise anatomic depiction, aiding accurate planning of ablation.¹

The role of CT in cardiac arrhythmias, the focus of this chapter, has benefited from the use of electrocardiogram (ECG)-gated image acquisition, a rapid increase in gantry rotational velocity, and an increasing number and resolution of available detector rows. Recently available 64-slice scanners provide high-resolution volume and improved image quality by effectively freezing the heart’s motion during acquisition and allowing scans to be performed more quickly or with fewer gantry rotations than with earlier scanners. This, in turn, has led to a reduction in errors associated with image reconstruction and segmentation, and more precise information on the position and movement of the cardiac structures throughout the cardiac cycle.

This chapter addresses the basic principles of CT imaging within the particular constraints of cardiac imaging relating to cardiac arrhythmias. Although CT imaging has been used to identify many structures in the heart, given the current interest in image-guided therapy for AF, CT imaging of the left atrium and its role in AF ablation is the main focus of this chapter.

Most medical images are in a digital format and are made up of an array of small square or rectangular elements called pixels.² Each pixel has associated image intensity. This provides the coordinate system of the image, and an element in the image can be assessed by its two-dimensional position within this array. For example, a typical CT slice is formed of 512 × 512 pixels, each corresponding to a portion of the cut through the patient measuring about 0.5 × 0.5 mm². The matrix and the pixel size are related to the display field of view (FOV). If, for example, the FOV is 25 cm, each pixel will be FOV/matrix size 25/512 = 0.48 mm². This dimension determines the limiting in-plane spatial resolution of the image. The two-dimensional axial slices are then stacked together to form a three-dimensional volume. Each pixel corresponds to a small volume element called a voxel. The height of the voxel is determined by the slice thickness. If the axial slice thickness in the above example is 1.5 mm, the voxel size would be 0.48 × 0.48 × 1.5 mm³.

Computed Tomography Imaging

In a CT imaging system configuration, an x-ray projects a fan-shaped beam that is collimated to lie within an x-y plane of a Cartesian coordinate system and generally referred to as the imaging plane.^1,² Thus, during CT imaging, the anatomy of interest passes through this imaging plane, and the image data of interest are acquired and reconstructed. This acquisition is typically accomplished by obtaining different views as the x-ray source and detectors rotate around the anatomy or volume. Reconstruction of these data generates a two-dimensional array of quantized grayscale values or pixels. Pixel values are a measure of the x-ray attenuation in Hounsfield units (HU), where the HU = 1000 – (4µ/µw – 1), µ being the average linear attenuation coefficient of the volume element represented by the pixel and µw the linear coefficient of water for the effective energy at the beam exiting the patient. Thus, water has an HU number of 0, and a region with a CT number of 100 HU has a linear attenuation coefficient that is 1% greater than the linear attenuation coefficient of water.

In the most commonly used ECG-gated helical CT acquisition, the x-ray tube (and detector array) rotates continuously about the patient collecting the data, while the table (with the patient) is moved at a constant speed through the imaging plane. The helical CT thus collects a continuous sequence of consecutive axial images from a volume of the patient’s anatomy. Faster scanning, as currently done with 64-slice CT scanners, allows collection of a large volume of data in short periods (<10 seconds). The factors selected in scanning a patient include slice width, FOV, gantry rotation speed, and volume of coverage, as well as basic contrast injection protocols. Scan time is calculated from the scan volume (total distance traveled by the table) divided by table speed. Also important is the pitch, which is defined as a ratio of the distance the table moves per 360-degree rotation. A pitch of 1 means the patient table moves a distance of one slice. Similarly, a pitch of 0.2 means the gantry rotates five times as the table moves a distance equal to the collimator width.

Cardiac motion caused by heartbeat, respiration, and patient movement while on the table can produce artifacts that appear as blurring in the reconstructed image. Such blurring effects may make diagnosis difficult. The use of a short scan time, as can be done with current scanners, can prevent or minimize these artifacts because of the speed of acquisition. To avoid respiration artifacts, scanning is performed during the breath held in inspiration or expiration. Because of the short scan time, currently available faster scanners allow images to be acquired in expiration. The acquired data are synchronized with the collection of the ECG (QRS) signal. The ECG signal is recorded in parallel with the CT through a noninvasive monitoring device connected to the patient. The data acquired during consecutive cardiac time intervals can then be combined to produce an image of the heart at the same phase of the cardiac cycle. Retrospective gating allows alignment of images during any phase of the cardiac cycle because of continuous helical acquisition. Usually, during retrospective gating, an approximately 75% phase location is used for patients in sinus rhythm as it yields the best image quality because of diastole. During AF, because of short R-R intervals, an approximately 45% cardiac phase usually gives best results.

Once the image is acquired, it is stored in a proprietary format. The data can then be exported from the scanner. As images obtained by scanners of one manufacturer may need to be imported to that of another or to different viewing screens, a medical image standard known as DICOM (Digital Information and Communications in Medicine) has been devised and is widely used. This allows data to be exchanged between scanners and viewing consoles.

The American Radiology Society’s convention is to display axial images with the right side of the patient at the left and the posterior side at the bottom when viewed on the computer screen at the workstation.

Image Segmentation

Following scanning, cardiac images are generated by postprocessing of one phase of axial image datasets. The cardiac chamber volumes are based on the boundary between the contrast-enhanced blood pool, which is of bright appearance because of the contrast, and the endocardium, which is not contrast enhanced. This allows for clear differentiation of the lumen and the myocardial wall. The process of dividing images into different regions to visualize areas of interest is called segmentation.^3,⁴ Image segmentation methods can be grouped into thresholding, boundary detection, and region identification.

Thresholding is the simplest but most effective segmentation method. During thresholding, pixels with intensities below a threshold value are assigned a certain class and the remaining pixels a different class. Connecting adjacent pixels of the same class forms regions. Thus, boundary extraction methods use information about intensity differences between adjacent regions to separate the regions from each other. Region identification techniques then form regions by combining pixels of similar properties. The simplest region-identifying technique could start with at least one seed (a starting point) per region. Neighbors of the seed are visited, and the neighbors that satisfy the predicate (a simple predicate compares the intensity values of the pixel to the average intensity value of the region) are added to the region. Segmentation, as relevant to AF ablation, for example, can be used to identify the anatomy of the LA and its vasculature that can be viewed separately from the remaining chambers of the heart and surrounding anatomy such as the lung.

Three-dimensional endocardial views, navigator views, and various measurements can be obtained from the imaged and segmented data (Figures 17-1 and 17-2). Cut planes can be used to remove a portion of the surface. The resultant model can show endocardial surface and pulmonary veins (PVs) as if from inside the chamber. Cutting away the anterior surface, for example, gives a good view of the posterior LA endocardium. Left and right anterior views can provide excellent delineation of right and left PVs, respectively. The navigator view shows the LA from the perspective of a virtual endoscope. Several measurements such as the LA volume (Figure 17-3), PV ostium (Figure 17-4), mitral isthmus and distance between the superior PVs (Figure 17-5), and relationship of the esophagus to the LA (Figure 17-6) can be made due to their relevance to some of the linear and other lesions performed in some AF procedures in these areas. For LA dimension measurement, a series of three lines or axes are drawn in the LA to measure chamber dimensions and to serve as the basis for creating a coordinate system in the LA. First, a line is drawn near the posterior portion of the LA connecting the junction point of the right superior and inferior veins to the junction point of the left superior and inferior veins. This creates the x-axis. Next, a line is drawn through the mid-portion of the x-axis, forming the y-axis. Finally, drawing a line through the intersecting points of the x– and y-axes creates a z-axis of the coordinate system. The length of these lines serves as the LA dimensions. In the case of measuring mitral isthmus dimensions, for example, markers can be placed on the three-dimensional volume at the inferolateral aspect of the left inferior PV, the posterolateral aspect of the mitral annulus, and midway between these two points help define the optimal line. Then, a cut plane that passes through these points is defined. Finally, the distance along this surface reaching from the left inferior PV to the mitral annulus is measured. PV ostial dimensions are measured using a standard double-oblique approach. An initial oblique cut is made along the shaft or lumen of the vessel. A second oblique cut is then made and positioned orthogonally to the first cut plane. The long and short axes are then measured. For LA esophageal measurement, axial slices are scanned. A point is deposited in the center of the esophageal lumen. An oblique cut plane is then created through this point, orthogonal to the posterior LA and the esophagus. Four measurements are made of the distance between the posterior LA endocardium and the esophagus.

FIGURE 17-1 Left atrial three-dimensional image. Left, A posteroanterior view of the segmented left atrium from the computed tomography image. Right, Endocardial views of the left atrium. Top right, A right lateral view, with the medial side of the left atrium cut away, revealing the left superior pulmonary vein (LSPV), left inferior pulmonary vein (LIPV), the left atrial appendage (LAA), as well as the mitral valve (MV). Bottom right, Left lateral portion of the left atrium cut away, revealing the right superior pulmonary vein (RSPV) and right inferior pulmonary vein (RIPV).

FIGURE 17-2 Navigator view of the left atrium. Polar view of the left atrium depicts the entire left atrium. LAA, Left atrial appendage; LIPV, left inferior pulmonary vein; LSPV, left superior pulmonary vein; MV, mitral valve; RIPV, right inferior pulmonary vein; RSPV, right superior pulmonary vein.

FIGURE 17-3 Left atrial measurement. Methods of measuring transverse, lateral, and anteroposterior length of the left atrium are shown. Localization of points over the three-dimensional (3D) computed tomography model made to create a localization line over the axial datasets is depicted. The image has been rotated to visualize all the highlighted points simultaneously.

FIGURE 17-4 Right superior (left) and left superior pulmonary vein ostial (right) measurements from axial datasets. The first oblique cut creates the planes (not shown) in the figure. The second oblique shown in the left and right panel results in the cross-sections of the pulmonary veins at the ostium.

FIGURE 17-5 Measurements of strategic locations used for ablation. A, The roofline connecting the superior pulmonary veins. B, The definition of the left inferior pulmonary vein (LIPV) to mitral valve annulus (MVA) line, mitral isthmus line. B is an oblique cut showing the MVA and LIPV and illustrates how the line, which follows posteriorly between the LIPV and the MVA, is drawn and measured.

FIGURE 17-6 Relationship of the esophagus to the posterior left atrium. Left lateral three-dimensional (3D) computed tomography images of the left atrium are shown. The esophagus is seen as a translucent structure near the left atrial posterior wall over the left pulmonary veins. Measurements at four different locations between the left atrium and the esophagus are shown.

Imaging of the Atrium and Pulmonary Veins

A thorough understanding of the morphologic characteristics of the LA and PVs in detail will not only help achieve a more efficient ablation but also prevent procedure-related complications such as PV stenosis and others by delineating the relationship of the LA to surrounding structures such as the esophagus and by helping to choose the right tools for mapping and ablation. A survey given to task force members for the AF ablation consensus document revealed that approximately two thirds of centers are routinely obtaining MRI or CT scans in patients scheduled to have an AF ablation.⁵

Detailed imaging studies have shown that anywhere from 65% to 80% of patients have four PVs, and some have left common and right middle PVs as well (Figure 17-7).⁶ Part of the main trunk of the right superior PV passes immediately behind the right superior vena cava (SVC) junction. It has also been shown that the right superior PV trunk branches out significantly sooner than do the left PVs. The right inferior PV arises inferiorly and laterally to the right superior PV. It divides almost immediately, within 5 to 10 mm, into superior and inferior branches. The distance between the right superior and right inferior PVs across the canal ridge varies from 2 to 8 mm. In 18% to 29% of cases, a supernumerary right PV may arise independently on the right side.⁷

FIGURE 17-7 Three-dimensional and endocardial left atrial image reconstruction. Representative examples of three-dimensional models of three different pulmonary vein morphologies, along with endocardial views, are depicted. The standard four pulmonary veins (top left), a common pulmonary vein (top middle), and an additional right middle vein (right) are shown. Bottom, The respective endocardial views, along with the mitral valve (MV). Left, The location of the circumflex artery (Cx). LAA, Left atrial appendage; LIPV, left inferior pulmonary vein; LPV, left pulmonary vein; LSPV, left superior pulmonary vein; RIPV, right inferior pulmonary vein; RMPV, right middle pulmonary vein; RSPV, right superior pulmonary vein.

(From Sra J, Krum D, Okerlund D, et al: Endocardial imaging of the left atrium in patients with atrial fibrillation, J Cardiovasc Electrophysiol Imag 15:247, 2004.)

The left superior PV lies superiorly and posteriorly to the LA appendage. It enters the LA in a more vertical direction. It usually has multiple branches, which ordinarily arise 10 to 20 mm from their base. The left inferior PV enters the LA more horizontally from a posterolateral position and branches almost immediately. A common antrum of the left superior and inferior PVs is seen in 3% to 30% of patients. Some studies have suggested this number may be even higher. In a series of more than 500 CT scans done at the author’s institution, in addition to the left common and right middle PVs, other unusual anatomies, including one common right PV, three PVs on the right and left, and one common ostium of the left inferior PVs, were seen.⁸ Examples are depicted in Figures 17-8 and 17-9.

FIGURE 17-8 Different pulmonary anatomies. A, Common right pulmonary vein (RPV). Endocardial view shows the common ostium (COs). B, Reconstructed three-dimensional model of the left atrium and pulmonary veins (PVs). In addition to superior and inferior PVs, three-dimensional model shows middle right and left pulmonary vein. Endocardial views show right and left PVs with additional pulmonary veins as seen on the three-dimensional model. LSPV, Left superior PV; LIPV, left inferior PV; LAA, left atrial appendage.

(From Sra J, Akhtar M: Mapping techniques for atrial fibrillation ablation, Curr Probl Cardiol 12:667–718, 2007. )

FIGURE 17-9 Unusual pulmonary vein anatomies. A and B, An unusual anatomic subtype of pulmonary vein anatomy. The left and right upper pulmonary veins (LSPV, RSPV) branch off from the left atrium (LA) in typical fashion; however, the two inferior pulmonary veins (LIPV, RIPV) converge near the midline of the posterior left atrium, forming a large common ostium (Common Os). C and D, Absent LIPV. Endocardial view is also depicted. LAA, Left atrial appendage.

(From Sra J, Malloy A, Shah H, et al: Common ostium of the inferior pulmonary veins in a patient undergoing left atrial ablation for atrial fibrillation, J Interv Card Electrophysiol 15:203, 2006; and Arora V, Nangia V, Krum D, et al: Absent left inferior pulmonary vein in a patient undergoing atrial fibrillation ablation. EP images from cell to bedside, J Cardiovasc Electrophysiol 16:924–925, 2005.)

PV ostia are ellipsoid, with a longer superoinferior dimension, and funnel-shaped ostia are frequently noted in AF patients. PVs are larger in patients with AF versus those without AF, in men versus women, and in persistent versus paroxysmal patterns.⁹ The understanding of these anatomic relationships is essential for accomplishing safe access to the LA using the trans-septal puncture, for placement of appropriate mapping tools such as a circular mapping catheter or multi-electrode basket catheter as well as Cryo balloons, and for application of energy around or outside the PV ostia. The variability of PV morphologies could substantially influence the success rate of catheter ablation if the variant veins are inadequately treated. Multiple ramifications and early branching observed in right inferior PVs possibly account for the lower incidence of focal origin of AF from this vein. These anatomic variations are important in planning catheter ablation of AF (Figure 17-10). Localization of the true LA PV, the LA appendage, and the ridge between the PV and the LA appendage in these patients can be more accurate with the assistance of three-dimensional CT images before mapping and ablation procedures.¹⁰

FIGURE 17-10 Endocardial computed tomography image showing pulmonary veins and different techniques of pulmonary vein isolation. A multi-electrode catheter is positioned in the upper pulmonary vein. Bottom vein shows circumferential isolation techniques used during atrial fibrillation ablation.

Left Atrial Registration

To improve intraprocedural guidance using current imaging techniques for ablation, cardiac image registration is currently under investigation and is in clinical use for AF ablation. Table 17-1 depicts some of the studies published in this regard. Cardiac image registration, which involves integration of two images in the context of the LA, is intermodal with the acquired image and the real-time reference image residing in different image spaces and involves optimization, where one image space is transformed into the other. Unlike rigid body registration, cardiac image registration is unique and challenging because of cardiac motion during the cardiac cycle and respiration motion. Registration algorithms involve the optimization of a cost function by the choice of a transformation, which transforms one image space into the other. The transformation can be either linear or nonlinear. Linear transformations are shape preserving and are composed solely of rotations, translations, and isotropic scaling. Nonlinear transformations may deform both the shapes and sizes of images. A linear transformation between three-dimensional spaces is defined by six parameters (or degrees of freedom), where two positions of a rigid body can always be related to one another in terms of three translations and three rotations. As the voxel sizes in each image may not be similar for calibration purposes, three extra degrees of freedom, equating to scaling in each direction, are needed. A simplified rigid body registration involves translation, scaling, and rotation, where the centroid, or the center part, is aligned in each image. Subsequently, scaling is performed to calibrate both images. This is followed by rotation to align the fiducial points. A nonlinear transformation will require more degrees of freedom.

TABLE 17-1 Registration Techniques and Errors

Many steps have been taken recently to develop methods of integrating three-dimensional structural details from acquired cardiac images with the real-time view of the interventional systems. The main modalities for catheter viewing, mapping systems, fluoroscopy, and ultrasound have been used in these techniques.¹¹ The following section describes some recent advances in the registration of acquired, structurally revealing three-dimensional images with real-time images.

Registration Using Anatomic Mapping Systems

Anatomic mapping systems provide the three-dimensional position of a navigational catheter within the cardiac chamber of interest and, in some instances, can also be used to construct three-dimensional maps of the cardiac chamber. Magellan and CARTO (Biosense Webster Inc., Diamond Bar, CA) use the electromagnetic position of the catheter tip, based on an electromagnetic locator pad, which is placed below the patient, and a reference catheter is placed at a fixed external (usually posterior) location. LocaLisa (Medtronic Inc., Minneapolis, MN) and NavX (St Jude Medical Inc., St Paul, MN) catheters are used to register the image. Previous results of this method have indicated that registration of the left ventricle alone results in inaccurate alignment. Inclusion of the aorta in the registration process rectifies this error. A clinical application of this technique that uses the CARTO system (CartoMerge), the multi-electrode catheter, and the NavX with either CT or MRI is now available.^11,¹² A combination of landmark and surface registrations is used to register CT with the CARTO system. Initially, several landmarks, usually three, are manually chosen and annotated. After this, the reconstructed three-dimensional image of the LA using CT or MRI is superimposed on the electroanatomic map created by the CARTO mapping system. Figure 17-11 depicts an example of CARTO and CT registration using landmark (fiducial point) and surface registrations.

FIGURE 17-11 Registration using a multi-electrode catheter and CARTO (Biosense Webster) with computed tomography (CT). A, Registration of the left atrium and multi-electrode balloon. Three-dimensional (3D) image of the left atrium and pulmonary veins (PVs) is visualized in the posteroanterior view using CT (left). The location of the buried electrodes (red circles) are deposited at the time of segmentation. The 3D model is imported using a wireframe model (middle) and registered with the noncontact mapping system (right) using the process detailed in the text. A sinus beat is identified in the 3D registered model. LAA, Left atrial appendage; LIPV, left inferior pulmonary vein; LMPV, left middle pulmonary vein; LSPV, left superior pulmonary vein; RIPV, right inferior pulmonary vein; RMPV, right middle pulmonary vein; RSPV, right superior pulmonary vein. Bottom, Registration using the CARTO system. B1, Landmark pairs (highlighted with darker circles) at the 6 o’clock mitral annulus (MA) position and the junctions of the LA and right superior pulmonary vein (RS), right inferior pulmonary vein (RI), and left inferior pulmonary vein (LI) were annotated on the 3D CT left atrial surface reconstruction (upper image, shown as wire frame) and the left atrial electroanatomic map (lower image, shown as solid shell) with tubes representing pulmonary veins. B2, After landmark registration, the 3D left atrial surface reconstruction was superimposed on the electroanatomic map (shown as mesh). Note: The misalignment of the LS, LI, and RS between the two image datasets is indicated by the yellow or red color of the PV points (arrows) sampled in those PVs. B3, After surface registration was executed, the PV alignment between the two image datasets was significantly improved, indicated by the PV points. The electroanatomic map points indicate their distance of <5 mm, 5 to 10 mm, and >10 mm, respectively, from the registered CT reconstruction surface.

(From Sra J, Krum D, Hare J, et al: Feasibility and validation of registration of three-dimensional left atrial models derived from computed tomography with a noncontact cardiac mapping system, Heart Rhythm 2:55–63, 2005; and Dong J, Dickfeld T, Dalal D, et al: Initial experience in the use of integrated electroanatomic mapping with three-dimensional MR/CT images to guide catheter ablation of atrial fibrillation, J Cardiovasc Electrophysiol 17:459–466, 2006.)

X-Ray Registration

Registration with fluoroscopy can be performed in the exported three-dimensional model by using a transformation to align the coronary sinus catheter seen on fluoroscopy with the SVC and the coronary sinus in the exported three-dimensional CT model. The author’s center has recently described an implementation of a semi-automated three-dimensional/two-dimensional CT-fluoroscopy registration strategy.¹³ The accuracy of this system was found to be within 1.4 mm in phantom studies. This strategy was also assessed in patients undergoing AF ablation. Twenty consecutive patients underwent ECG-gated, contrast-enhanced CT scanning. The LA and the PVs were segmented using the semi-automated method described before. The segmented images of the LA PV were then registered in real time on acquired digital cine or fluoroscopic images by superimposing the coronary sinus catheter, as seen on the cine image, over the SVC and coronary sinus as described before. Accurate registration was confirmed by PV angiography as well as by the position and recordings on a 64-pole basket catheter located in the PVs (Figure 17-12). The author’s center has continued to use this strategy for PV isolation and LA linear ablation procedures on patients with AF. In a recent study, 50 patients with AF undergoing ablation were randomized to standard ablation techniques or ablation guided by CT-fluoroscopy registered images. Fluoroscopy time and total procedure time were significantly reduced in the CT-fluoroscopy group, and a trend toward better outcomes was observed. The complex translational, rotational, and conformational changes that occur with cardiac and respiratory motions will introduce error into the registration process when, as in three-dimensional/two-dimensional registration, a static image is “aligned” with the real-time fluoroscopic image. Attempts at gating the registration process so that image integration occurred during the same phase of the cardiac cycle were successfully implemented in experiments conducted in the author’s laboratory; thus, fluoroscopic images can be taken at the same phase as the segmented CT images, that is, during diastole (see Figure 7-12). Synchronizing registration to the respiratory cycle during expiration or inspiration, the phase during which CT imaging has been done, could potentially eliminate movement of the LA during respiration.

FIGURE 17-12 Electrocardiogram and respiration-gated registration. Fluoroscopic image is taken during the same cardiac cycle phase (75%, green arrows), and respiration gated (during inspiration) as computed tomography imaging. Accurate registration is seen. System also allows marking of the ablation site (red circles).

(From Sra J: Cardiac image registration, JAFIB J Atrial Fibrillation 1:148–160, 2008.)

Other Applications in Cardiac Arrhythmias

The ability to distinguish the dysfunctional but viable myocardium from nonviable tissue may have important prognostic implications after myocardial infarction (MI). A study by Lardo et al validated the accuracy of contrast-enhanced multi-detector CT (MDCT) for quantifying myocardial necrosis, microvascular obstruction, and chronic scar after occlusion or reperfusion MI.¹⁴ In this study, 10 dogs and 7 pigs underwent balloon occlusion of the left anterior descending (LAD) coronary artery followed by reperfusion. Contrast-enhanced (Visipaque, 150 mL, 325 mg/mL) MDCT (0.5 mm, 32-slice) was performed before occlusion and 90 minutes (canine) or 8 weeks (porcine) after reperfusion. MDCT images were analyzed to define infarct size and extent and microvascular obstruction and compared with postmortem myocardial staining (triphenyltetrazolium chloride) and microsphere blood flow measurements. Acute and chronic infarcts by MDCT were characterized by hyper-enhancement, whereas regions of microvascular obstruction were characterized by hypo-enhancement. MDCT infarct volume compared well with triphenyltetrazolium chloride staining (acute infarcts, 21.1% ± 7.2% vs. 20.4% ± 7.4%; mean difference, 0.7%; chronic infarcts, 4.15% ± 1.93% vs. 4.92% ± 2.06%; mean difference 0.76%) and accurately reflected the morphology and the transmural extent of injury in all of the animals (Figure 17-13). Peak hyper-enhancement of infarcted regions occurred 5 minutes after contrast injection. MDCT-derived regions of microvascular obstruction were also identified accurately in acute studies and correlated with reduced-flow regions as measured by microsphere blood flow. This study thus suggested that the spatial extent of acute and healed MI could be determined and quantified accurately with contrast-enhanced MDCT. This feature, combined with existing high-resolution MDCT coronary angiography, may have important implications for the comprehensive assessment of cardiovascular disease.

FIGURE 17-13 Multi-detector computed tomography (MDCT) and histopathologic staining comparison of infarct morphology. A, Reconstructed short-axis MDCT slice 5 minutes after contrast injection demonstrating a large anterolateral infarct (hyper-enhanced region) with discrete endocardial regions of microvascular obstruction (four arrows). B, Triphenyltetrazolium chloride (TTC)–stained slice. C, Thioflavin S and TTC staining of the same slice, which confirms the size and location of microvascular obstruction regions. D, Quantitative MDCT and TTC measurements of infarct size yielded good agreement, with points distributed around the line of identity. E, Mean difference of 0.7% by Bland-Altman analysis.

(From Lardo AC, Cordeiro AS, Silva C, et al: Contrast-enhanced multidetector computed tomography viability imaging after myocardial infarction characterization of myocyte death, microvascular obstruction, and chronic scar, Circulation 113:394–404, 2006.)

Other studies have shown the role of CT in imaging coronary sinus and left ventricular anatomy and function to define sites for biventricular pacing in patients with congestive heart failure (Figures 17-14 to 17-17). CT imaging has also been used to perform left ventricular functional analysis using post-processing algorithms or detailed left ventricular motion analysis, thus enabling identification of sites where left ventricular pacing should improve efficacy of biventricular pacing.¹⁵

FIGURE 17-14 Three-dimensional images of the left ventricle (LV) and coronary sinus (CS). A, Endocardial images of the CS also can be obtained using post-processing segmentation algorithms. A three-dimensional image of the LV and the CS with its branches in a patient obtained using CT (16 slice, light speed ultra). B, Post-processing segmentation algorithms used to show three-dimensional images of the CS alone. C and D, Endocardial views of the CS at two different locations. These imaging techniques could help create a roadmap for CS lead placement in patients undergoing cardiac resynchronization therapy.

(From Sra J, Krum D, Okerlund D, et al: Three-dimensional and endocardial imaging of the coronary sinus for cardiac resynchronization therapy, J Cardiovasc Electrophysiol 15:1109, 2004.)

FIGURE 17-15 Ventricular wall motion using three-dimensional imaging. A, An example during systole and diastole, with the endocardium represented by the red mesh and the epicardium represented by the green mesh. This information can be processed to determine optimal areas for left ventricular pacing, such as the site that is the last to attain maximum displacement and the last to reach maximum velocity. Following scanning and segmentation, each axial slice (thickness, 1.25 mm) was divided into 100 chords, representing the full circumference of the axial slice. The wall motion of the endocardium at each of these chords, expressed as a displacement from end diastole was plotted throughout the cardiac cycle. Data from these chords were averaged over six areas (anterior [A], anterolateral [AL], lateral [L], posterior [P], posterolateral [PL], septal [S], anteroseptal [AS]) in 5% increments throughout the cardiac cycle. B, A plot of wall motion versus time for a slice midway between the apex and the base of the heart. LV wall motion and velocity also were determined from these data. C and D, A plot of wall motion versus time and velocity versus time for the same axial slice. By identifying areas of interest, three-dimensional cardiac imaging creates a roadmap that may help optimize cardiac resynchronization therapy.

(From Sra J, Krum D, Okerlund D, et al: Ventricular wall motion using three-dimensional imaging, J Cardiovasc Electrophysiol 15:1110, 2004.)

FIGURE 17-16 Ventricular wall motion during left bundle branch block and normal QRS. An example during systole and diastole is shown with the endocardium represented by the red mesh and the epicardium represented by the green mesh. Left, Normal QRS. Right, Ventricular wall motion during left bundle branch block. As opposed to normal QRS, asynchronous contraction of the ventricle is seen in left bundle branch block.

FIGURE 17-17 Right and left ventricular wall motion during left bundle branch block in the patient with left bundle branch block, as shown in Figure 17-16. The right ventricle is shown in yellow. During systole, asynchronous contraction of the left ventricle is seen. Post, Posterior; Lat, lateral; Ant, anterior; Sept, septal.

Conclusion

Recent advances in imaging, especially in CT imaging, may help define anatomic structures associated with cardiac arrhythmias. This is particularly true for complex three-dimensional anatomy such as the LA. Image registration of these anatomic structures further provides a means for physicians to incorporate into a single view the varied information captured by CT, which can then be used for interventional planning and treatment of complex arrhythmias such as AF. Significant work is being done to improve imaging quality and function assessment, which should further refine the imaging of structures such as the coronary sinus and the left ventricle and the identification of myocardial scar. This should enhance efficacy of procedures such as biventricular pacing.

Acknowledgments

The authors thank Brian Miller and Brian Schurrer for assistance in the preparation of illustrations and Barbara Danek and Joe Grundle for editing the manuscript.

B. Cardiac Magnetic Resonance Imaging

Introduction

Improvements in understanding the arrhythmia mechanisms and the development of advanced three-dimensional mapping tools have helped create novel catheter-based ablation strategies over the past 10 to 15 years. As many of these approaches are anatomically based, cardiac imaging has been increasingly used to assist in procedural planning, guidance, and follow-up.

Three-dimensional mapping systems became available in the mid-1990s, allowing the real-time display of a catheter tip within a mathematically reconstructed cardiac chamber.¹⁶ However, they lacked the anatomic detail desired for complex ablation procedures. Therefore, cardiac imaging was increasingly used in patients undergoing complex ablations such as AF ablations. Today, pre-procedural MRI or CT is used to characterize the individual anatomies of the LA and the PV and has become standard-of-care in many arrhythmia centers. This allows for a tailored, patient-specific ablation approach. Detailed anatomic information gained from cardiac imaging is being made available during the ablation procedure by integrating these images with three-dimensional mapping systems. In addition, imaging is used in the detection of procedural complications and postprocedural patient care.

Cardiac imaging is also increasingly used for ventricular tachycardia (VT) ablations. MRI is well validated to provide detailed information about the cardiac anatomy, the myocardial scar, or both, which are usually the targets for substrate-guided VT ablations. Current software allows the export of detailed imaging information into three-dimensional mapping systems to provide anatomic catheter guidance for patients with recurrent episodes of VT.

An emerging application of cardiac imaging is the visualization of atrial or ventricular ablation lesions using MRI. This has the potential to assess the discrepancies between intended versus actual ablation lines, delineate gaps in ablation sets, and provide a road map for additional and repeat ablation procedures.

A rapidly evolving field is real-time MRI for electrophysiology procedures, which is currently being evaluated in animal and preliminary human studies. Catheter navigation and ablation can be guided under real-time MRI visualization, which identifies the precise location of the ablation catheter tip in relation to the surrounding anatomy. This novel imaging paradigm has the potential for direct confirmation of the exact anatomic catheter location, verification of catheter-tissue contact, lesion visualization, and early detection of complications.

Current Technology of Image Integration

The ability to visualize the true three-dimensional anatomy from cardiac imaging fueled the integration of these datasets with clinical three-dimensional mapping systems to provide the electrophysiologist with patient-specific cardiac anatomy during the ablation procedure. Before the era of three-dimensional mapping, catheter navigation was guided by fluoroscopy with its known limitations with regard to visualization of soft tissue, catheter contact, and patient exposure to ionizing radiation. In the late 1990s, three-dimensional mapping systems that significantly reduced fluoroscopy time up to 82% for atrial flutter and 49% for AF became available.^17,¹⁸

The two most commonly used three-dimensional mapping systems are Biosense Webster’s CartoMerge and St. Jude Medical’s Ensite NavX navigation system. The former uses electromagnetic fields to locate an endocardial catheter, whereas the latter uses electrical signals transmitted through the patient’s body. Image integration refers to a process in which the detailed three-dimensional anatomy from MRI or CT is extracted and superimposed on the endocardial shell created from catheter recordings (Figure 17-18). To create the endocardial shell with a three-dimensional mapping system, a roving catheter is moved sequentially along the endocardial surface, and location points are acquired. On the basis of the collected points, an electroanatomic shell which approximates the actual cardiac anatomy is mathematically reconstructed. An accurate registration (alignment of the MRI-derived dataset with the catheter-created shell) is critical for accurate anatomic catheter placement. Registration is performed by selecting several corresponding, well-defined anatomic landmarks such as the PV ostia and the posterior LA wall on the catheter-derived shell and the MRI-derived three-dimensional model. Depending on the mapping system, additional proprietary algorithms can adjust the three-dimensional model position to achieve the minimal average registration error between the shell and three-dimensional model points (surface registration) or will adjust points of the catheter-derived map to fit into the three-dimensional anatomic model (dynamic registration).^12,¹⁹ The resulting registration accuracies have been reported in the range of 1 to 4 mm.^12,¹⁹