Indication

This method, also known as a fast low-angle shot (FLASH), or fast-field echo (FFE), is the most extensively studied for assessing ventricular wall motion, volumes, mass, and EF. It is particularly useful for visualizing turbulent flow associated with valvular stenosis and regurgitation.

Limitation

In patients with severe ventricular dysfunction, slow blood flow along the endocardial surface reduces one’s ability to visualize low contrast in the LV myocardium.

Description

Using spoiled gradient-echo techniques, the movement of blood through the LV cavity provides bright contrast against the gray appearance of the myocardium.

Limitation

This technique underestimates turbulent flow associated with intracardiac shunts or valvular heart disease (Table 16-1).

Description

SSFP imaging exhibits high signal-to-noise ratio (SNR) and high contrast-to-noise ratio (CNR) with the blood-myocardial interface. With SSFP, cine MR images depict the endocardial surface, regardless of blood flow velocity.^7,8 This technique can be used with brief repetition times repetition times (TR) (<3 ms), leading to short scan times with high temporal resolution. Studies comparing FLASH and SSFP imaging have demonstrated significant improvements with SSFP in the blood-myocardial interface in patients with decreased LV ejection fraction.⁹ With short echo times of less than 1.5 ms, the blood pool within the cavity appears more uniform and image quality is not hampered by turbulent flow.

Indication

Echo-planar imaging (EPI) is useful for obtaining a qualitative assessments of LV wall motion in patients unable to perform breath-holding.

Limitation

Because of low spatial and temporal resolution, this technique is not well suited for precise determination of volumes and EF.¹⁰

Description

Ultrafast EPI is a fast imaging sequence in which multiple echoes are acquired following each excitation that are then used to perform single-shot imaging. “Snapshots” of LV wall motion acquired during relatively short imaging times (e.g., 30-40 ms/slice) are produced with this technique by using advanced hardware to switch the gradients rapidly during scanning. This fast gradient-echo technique exhibits a relatively low SNR and more chemical shift artifacts compared with conventional gradient-echo methods.

Indication

The sensitivity-encoding (SENSE) technique obtains images during short periods of breath-holding.

Limitation

There is a loss of SNR in some images.

Description

This technique uses sensitivity information from multiple coil elements to correct for k-space undersampling in the post-Fourier domain. Each coil has a different sensitivity profile so that undersampled acquisitions may be unfolded that result in decreased scan time. Images produced with the SENSE technique exhibit consistent contrast, with preserved spatial resolution (see Table 16-1).

Indication

This is highly accurate for assessing LV midwall myocardial function.

Limitation

With this technique, it is difficult to measure LV function along the LV epicardial and endocardial surfaces.¹¹

Description

Myocardial motion can be tracked in one, two, or three dimensions using tissue tagging.^12,13 With tagging, dark saturation bands, or “markers,” are placed across the myocardium for the purpose of quantifying LV function. These markers are induced by prepulse sequences applied immediately after the R wave, usually in planes perpendicular to the imaging plane. Quantification of intramyocardial deformations can be accomplished by tracking the intersection points of the tagging lines to demonstrate myocardial rotation, contraction, relaxation, and strain (Fig. 16-1). The SPAMM (spatial modulation of magnetization) technique,¹⁴ uses two perpendicular sets of parallel lines that form a rectangular grid on the image that can be tracked throughout the cardiac cycle. These tagging lines move with the myocardium during the contraction and relaxation phases of the cardiac cycle (see Fig. 16-1).

FIGURE 16-1 Assessment of LV strain by MRI tagging. Shown are tagged MRI images of the mid–left ventricle in short axis at end-diastole (left) and end-systole (right). The deformation of the grid can be seen visually and can be used to determine the quantitative circumferential strain. For two-dimensional analysis, motion of the heart wall that transformed a hypothetical unit circle during diastole into an ellipse during systole corresponds to the directions of the eigenvectors of the transformation. Their lengths are the eigenvalues shown in the diagram below the images. The radial thickening or displacement is the difference between c′ length and c length.

Another tagging technique, complementary SPAMM (C-SPAMM), results in myocardial tag persistence throughout the entire cardiac cycle by using a negative tagging signal during the diastolic phases of image acquisition. The C-SPAMM technique allows for the study of both systolic and diastolic myocardial deformation resulting from improved tag contrast relative to the background myocardium. A disadvantage of C-SPAMM involves a longer image acquisition time. However, this technique is useful for patients with diastolic dysfunction or elevated heart rates (e.g., during dobutamine stress).

As soon as the images are acquired, there are three methods for extraction of motion data from tagged images: (1) tracking the dark tag lines as intensity minima; (2) eusing optical flow analysis; or (3) applying harmonic phase (HARP) determinates. All three of the tracking methods (Table 16-2) are highly accurate for assessing midwall myocardial function but exhibit some difficulty for discriminating tag intersection points near the epicardial and endocardial surfaces.¹¹

TABLE 16-2 Technique for Assessing Quantitative Analysis of Myocardial Motion

Global Left Ventricular Function

The two most common methods used to measure LV volume and EF are the area-length technique and the multislice Simpson’s rule technique.

Multislice Simpson’s Rule Method

Quantification of global function using multiple cine MRI provides measurements of volumes, EF, and mass that do not depend on assumed LV geometry. Using Simpson’s rule, stacked disks are assessed and the end-diastolic and end-systolic volumes are calculated by summing the volume of the blood pool in each slice. This method is highly accurate and reproducible, and has been validated in ex vivo and in vivo models. Figure 16-2 illustrates the method for obtaining standard views and Figure 16-3 shows some standard multislice and LV volume and mass calculations.

FIGURE 16-2 Method for obtained standard views, obtaining three short-axis (basal, mid, and apical) and three apical long axis views (long-axis, four-chamber, and two-chamber views). In all images, the myocardium is gray and the blood pool is white. The white dashed lines indicate the slice positions for obtaining the subsequent views demarcated by the arrows. Ao, aorta; LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle.

FIGURE 16-3 Method for obtaining standard multislice short-axis views from the four-chamber plane. The horizontal yellow lines indicate the slice positions for obtaining the short-axis view. Left, Each short axis slice is shown with the endocardial (red) and epicardial (green) borders used for the calculation of LV volume and LV mass.

(myocardial area = the difference in area between the endocardial and epicardial contour in the end-diastole, n = number of total slices. The specific gravity of the myocardial tissue = 1.05 g/dL.)

(EF = ejection fraction, EDV = end-diastolic volume, ESV = end-systolic volume.)

Left Ventricular Mass

Typically, LV mass is measured at end-diastole from multiple short-axis images. The epicardial and endocardial borders are traced in each slice, and papillary muscles and endocardial trabeculae may or may not be included in the LV mass calculation. Left ventricular mass is reported in grams and is calculated by multiplying the volume of the LV myocardium by the specific gravity of the myocardial tissue, 1.05 g/dL. The accuracy of LV mass has been established in autopsy studies of the heart and from studies in live human and animal subjects (see Fig. 16-3).

Regional Left Ventricular Function

Systolic Left Ventricular Function

Qualitative (Wall Motion)

With MRI, regional LV contractile function can be assessed qualitatively by visual inspection. Normally, LV free wall thickness increases more than 40% during systole. Regional hypokinesia is defined as systolic wall thickening less than 30%, akinesia is defined as systolic wall thickening less than 10%, and dyskinesia is defined as systolic outward movement or diastolic bulging of the ventricular wall.²⁴

Left ventricular wall motion is also visualized in orthogonal planes oriented parallel to the long axis of the heart. A combination of three short-axis cines spanning from the base to the apex are routinely acquired (Table 16-3) and apical planes are also acquired. These include the vertical long-axis (VLA) view, or four-chamber view, the horizontal long-axis (HLA) view, or two-chamber view, and the long axis or three-chamber view (Fig. 16-4; also see Fig. 16-2).²⁴

TABLE 16-3 Normal Values of Left Ventricular Function from Short-Axis View in Healthy Persons*

FIGURE 16-4 Cine MR images of the left ventricle are displayed in three short-axis (upper row) and three long-axis (lower row) views. Myocardial regions are divided into 17 segments, as identified: 1, basal anterior; 2, basal anteroseptum; 3, basal inferoseptum; 4, basal inferior; 5, basal inferolateral; 6, basal anterolateral; 7, mid-anterior; 8, mid-anteroseptum; 9, mid-inferoseptum; 10, mid-inferior; 11, mid-inferolateral; 12, mid-anterolateral; 13, apical anterior; 14, apical septum; 15, apical inferior; 16, apical lateral; and 17, apical cap. SAX, short axis.

Quantitative (Wall Thickening)

1 Modified center line method. Radial wall thickness is quantified by creating a radian emanating from the center of the left ventricle that courses perpendicularly across the endocardium and epicardium. The length between the endocardial and epicardial surface evenly spaced transmyocardial lines or chords identifies the local LV wall thickness. The difference between end-diastolic and end-systolic chord lengths represents local end-systolic wall thickening. This technique can overestimate wall thickness if the MRI slice is not acquired in a true short-axis plane.

2 Thickening or thinning assessed with tagging. Most tagged MRI data consist of two or three sets of two-dimensional images (see earlier). Tag deformation associated with ventricular motion and deformation provides information regarding circumferential and longitudinal shortening and lengthening, and radial thickening and thinning. Assessment of radial deformation can be compared within different myocardial segments to characterize regional wall motion (see Fig. 16-1).

Left Ventricular Strain

Left ventricular strain measurements reflect the direction and spatial variation of LV myocardial displacement without dependence on LV preload.^25,26 The LV myocardium has a complex architecture—subepicardial and subendocardial fibers are longitudinal, whereas mid-wall fibers are circumferential in direction. LV deformation is therefore composed of radial thickening, circumferential shortening and torsion, and longitudinal shortening. There are two coordinate systems in three orthogonal directions that are used to determine principle strains; radial strain is the myocardial deformation in the radial direction (R), longitudinal strain (L) is myocardial shortening from base to apex (negative value), and circumferential strain is the intramural circumferential shortening (C).²⁷ Shear strains represent the changes in angles between coordinate axes (Fig. 16-5).²⁸

FIGURE 16-5 Diagram showing the LV coordinate system includes radial (R), circumferential (C), and longitudinal (L) directions.

The circumferential-longitudinal shear angle describes the twisting motion of the heart and is closely related to torsion of the left ventricle. The twisting motion of the left ventricle stores potential elastic energy by straining the extracellular matrix, which is released during early diastole, and therefore contributes to early diastolic suction.²⁹ The contraction and relaxation of the spiraling myofibers cause the LV torsion and untwisting. Torsion is defined as the circumferential-longitudinal shear on the epicardial surface between two short-axis slices. The net ventricular torsion is the result of the subepicardial contraction and subendocardial counterbalancing contraction.²⁹ Both tagged imaging and DENSE techniques are well suited for tissue tracking for strain assessment. Several studies have reported the use of myocardial strain assessments to evaluate regional systolic and diastolic LV function in patients with myocardial ischemia, infarction, and postinfarction remodeling.^29,30

Diastolic Left Ventricular Function and Regional Relaxation

Parameters of diastolic function are determined from rates of inflow into the left ventricle or from the outward relaxation of the myocardium. Phase contrast velocity mapping can be used to quantify ventricular inflow in real time. This information is useful for evaluating the impact of respiratory motion on ventricular filling to demonstrate constrictive physiology (e.g., in constrictive pericarditis). Using the transmitral velocity to measure the mitral inflow pattern can also demonstrate LV filling patterns similar to these obtained with Doppler echocardiographic techniques (Fig. 16-6).

FIGURE 16-6 Mitral inflow velocities as seen by MRI phase contrast (left) and echo Doppler (right). Close agreement between the peak E and A velocities (about 55 and 35 cm/second, respectively) can be seen in both modalities

The assessment of outward movement of the LV myocardium can be acquired with tissue tagging techniques. By using myocardial tagging to evaluate the diastolic parameters, reverse myocardial strain and shear strain can be assessed at rest or during high-dose dobutamine stress to identify patients with flow-limiting epicardial coronary artery luminal narrowings.³¹

Global Right Ventricular Function

Cardiac MRI demonstrates high interstudy reproducibility for determining RV function (volumes and EF) in healthy subjects, patients with heart failure, and patients with hypertrophy.³² The intraobserver, interobserver, and interstudy variability was demonstrated to be 5% to 6% in normal volunteers.³³ Cardiac MRI has been used to follow patients with repaired tetralogy of Fallot. Davlouros and colleagues³⁴ have demonstrated pulmonary regurgitation and RV outflow tract aneurysm associated with RV dilation and a decrement in RV ejection fraction in patients after tetralogy of Fallot repair. Similar to the left ventricle, Simpson’s rule method is used to calculate RV volume and EF. Table 16-4 summarizes normal measures of RV function from various studies.

Regional Right Ventricular Function

There is no standard convention for segmenting the RV myocardium to determine regional RV function. In short-axis views, the RV free wall can be divided into four regions—the inflow tract, outflow tract, midventricular region, and apex. Arrhythmogenic right ventricular cardiomyopathy (ARVC) is a clinically important syndrome with regional RV motion abnormalities (akinesis or dyskinesis) on cine MRI that correspond to areas of adipose tissue replacement.³⁵ As shown in Figure 16-7, MRI is useful for identifying this condition.

FIGURE 16-7 Basal short-axis slices showing RV outflow tract dyskinesis in a patient with ARVC.

Although most myocardial tagging studies have been performed on the left ventricle, tagging of the right ventricle can be used to study RV myocardial deformation in patients with congenital heart diseases such as postoperative atrial switch procedures for transposition of the great arteries.³⁵ Complex RV geometry and the presence of a thin RV wall (normal < 6 mm) result in muscle tagging analyses being technically more demanding relative to the thicker walled left ventricle.

Right Ventricular Diastolic Function

Tricuspid transvalvular flow profiles may be used to assess RV diastolic filling patterns (E and A waves). Diastolic dysfunction within the walls of the right ventricle have been demonstrated using phase contrast velocity mapping in patients with hypertrophic cardiomyopathy.²⁵

KEY POINTS

Cardiac MRI yields fast, accurate, reproducible assessments of global and regional systolic and diastolic LV and RV function.

Multislice Simpson’s rule methods with cine white blood imaging techniques can be used to provide accurate measures of ventricular volume, mass, and ejection fraction.

A variety of quantitative analysis techniques incorporating tagging, phase contrast MRI, or displacement encoding with stimulated echoes (DENSE) can be used to provide precise measures of right or left ventricular systolic and diastolic function.