Strain and Strain Rate Imaging

6 Strain and Strain Rate Imaging

Thomas H. Marwick

Basic Principles

Assessment of Left Ventricular Function

• Assessment of left ventricular (LV) function is the most common indication for echocardiography.
• The most widely used parameter for this measurement is ejection fraction (EF). Although this measure is simple and intuitive, as well as being supported by a wealth of prognostic information, it has important limitations:
• Image quality dependence
• Geometric assumptions
• Load dependence
• Insensitivity to early disease
• There is an emphasis on moving from the diagnosis of late heart failure (advanced structural change, poor outcome) to early stage heart failure (limited structural change—more chance of reversal). Early myocardial disease is usually characterized by:
• Disturbances of longitudinal function
• Compensatory radial and circumferential function (and hence preserved EF)
• Recognition of longitudinal dysfunction is an important process in heart failure with normal EF.

Defining Strain

• Strain is a fundamental physical property of matter that reflects its deformation under an applied force.
• Strain can be expressed in any dimension—radial, longitudinal, or circumferential:
• Longitudinal deformation has been the most widely assessed.
• Speckle strain (see later discussion) can measure circumferential and radial strain more readily than tissue velocity techniques, but radial strain is difficult to measure by any method.
• Potentially, shear stresses between these planes may be identified.
• Strain may be measured as Lagrangian or natural strain:
• Lagrangian strain is defined on the basis of deformation from the original length (i.e., not subject to external forces): eL = {Lt − L(t=0)}/L(t=0)
• Natural strain is defined relative to a previous time instance but not original length: eN = ln {1 + et}
• Natural strain is more relevant to cardiac imaging.
• Lagrangian and natural strain are analogous at low strain (<10%), and one can be corrected to the other.
• In the tissue Doppler methodology, strain rate (SR) was derived from the gradient of velocity over a sampling distance, and strain obtained as the integral.
• Using speckle tracking, strain is derived from excursion of the speckles, and SR is derived from differentiation of the strain curve.
• By convention:
• Shortening is described as negative strain (the main longitudinal waveform).
• Lengthening is described by positive strain (the main radial component) (Fig. 6-1).
• Strain can be measured as:
• a magnitude (i.e., the nadir of the strain signal)
• a rate of change (SR)
• a timing parameter (time to peak strain)
• a mixture of magnitude and timing (SR, end-systolic strain)
• Potential magnitude parameters include:
• Peak systolic SR
• Peak isovolumic relaxation (IVR) SR (i.e., peak SR between aortic valve closure [AVC] and mitral valve opening [MVO])
• Peak early diastolic SR
• Peak systolic strain
• End-systolic strain
• Maximum strain
• Postsystolic thickening (maximum strain after AVC minus end-systolic strain)
• Potential timing parameters include:
• Time to relaxation (i.e., time to SR crossover at middle of cycle)
• Time to contraction (i.e., time to SR crossover at start of cycle)
• Time to peak SR
• Normal ranges of strain and SR have been described (Table 6-1). SR and strain may be expressed as curves relating to measures at one site (see Fig. 6-1) or in a parametric map (Fig. 6-2) where colors express magnitudes at various sites.
• There are minor differences between strain measurements provided by vendors. Nonetheless, strain and SR on all commercial systems have been extensively validated:
• Clinical validations have been performed on the basis of comparison with tagged magnetic resonance imaging (MRI) evidence of deformation. Unfortunately, this technique is imperfect.
• The most reliable validation is obtained in animal models using microcrystals to measure deformation.
image

Figure 6-1 Morphology of radial (above) and longitudinal (below) strain curves. Systolic strain (measured at aortic valve closure, AVC), may be exceeded by peak strain, and the difference may be measured in timing or strain magnitude (postsystolic shortening), usually measured as postsystolic index (PSI).

TABLE 6-1 NORMAL RANGES OF STRAIN AND SR HAVE BEEN DEFINED AT EACH PHASE OF THE CARDIAC CYCLE

image
image

Figure 6-2 Parametric map of apical long axis in which colors express velocity (above) and strain rate magnitudes and timing (below) starting at the posterior and progressing to the anteroseptal wall. The example shows the relative sensitivity of strain rate imaging (SRI) to small regions of wall motion abnormality. An infarct in the posterior wall (zone A) is not detected by tissue velocity imaging, which shows a red color, consistent with movement toward the transducer. This is the same as the anterior wall (zone B) and reflects tethering by the remainder of the wall, causing translational movement within the infarct area (upper panel). Because of the site specificity of SRI, tethering does not alter the deformation of the risk area (zone A), which is identified correctly as infarcted (note the blue color in systole), and contrasts with the anterior wall (zone A), which shows a gold color, consistent with thickening (zone B). As this is a Doppler-based technique, the apex (zone C) is not well evaluated by either method.

Benefits of Strain

• Deformation offers four main benefits in comparison with other quantitative parameters: sensitivity, site specificity (i.e., independence from tethering), homogeneity, and physiologic significance (correlation with contractility):
• The sensitivity of the deformation techniques to identify subtle gradations in function is an important attribute, especially for identifying early-stage disease. This is dependent on both having a narrow enough standard deviation to readily separate normal from abnormal tissue, as well as being specific for the location of the abnormal tissue (see Fig. 6-2).
• Site specificity implies that measurements obtained from within the sample volume truly reflect the performance of the tissue at this location rather than elsewhere in the wall. This is an inherent requirement for assessment of small areas of abnormal tissue, and techniques that are not site specific (e.g., tissue Doppler imaging or annular displacement) tend to average function along the length of the entire wall (see Fig. 6-2).
• The homogeneity of strain throughout the myocardium is relative rather than absolute, as there are minor variations of strain from apex to base. These variations are much less than the differences in tissue velocity within different parts of the heart (Fig. 6-3).
• The physiologic correlates of strain and SR have been sought in a number of experimental studies. Strain, which does not account for the time course of contraction, is analogous to regional EF. SR, a descriptor of the speed of contraction, is analogous to rate of change in pressure over time (dP/dt).
image

Figure 6-3 Regional variation of strain (A) and tissue velocity (B) in the septum. As the apex is fixed and the base moves away and toward this like a piston, there is a gradation of velocity from base (green) to apex (turquoise). In contrast, although there are minor regional variations of strain, strain values from apex and base are relatively uniform.

Disadvantages of Strain

• Measurement of myocardial deformation is a helpful but far from perfect technique for the assessment of regional or global LV function.
• Both tissue Doppler and speckle techniques are computationally complex.
• None of the methods is independent of user expertise, to the extent that a variety of artifacts need to be recognized and avoided, in order to prevent misleading results.
• The tissue velocity methodology is susceptible to artifact caused by angulation and works optimally when the direction of motion and the imaging access are as close as possible to parallel, and may be very misleading when angulation is major (Fig. 6-4).
• The speckle technique is very dependent on the quality.
image

Figure 6-4 Impact of angulation on strain measurement. In the apical view, the mid-septum (yellow) is closest to parallel with the ultrasound beam. As the incident angle increases towards 45 degrees, apparent strain measurement decreases (turquoise). In the apex, the incident angle increases to 90 degrees, so radial strain is measured, the strain direction being opposite to that of longitudinal strain.

Key Points

• LV EF is not an ideal measure of LV function because it depends on image quality, geometric assumptions, and loading conditions and is not sensitive for detection of early disease.
• Strain is defined as the change in myocardial length, relative either to original length or to length at a previous time, derived from excursion of speckles.
• SR is the rate of change in myocardial length (velocity), normalized to distance, and is the first derivative of the strain curve.
• SR and strain are displayed either as curves for one or more specific myocardial sites or as two-dimensional (2D)/three-dimensional (3D) maps of the left ventricle (LV) with color used to indicate magnitude over the entire LV.

Methodologies

• A variety of approaches to strain are now available from equipment vendors.
• They differ fundamentally according to whether tissue velocity or speckle is used as the primary modality (Fig. 6-5).
image

Figure 6-5 Myocardial imaging method. A, Tissue velocity imaging. B, Tissue velocity imaging (TVI)-based strain rate (SR) and strain. C, Speckle-based strain using velocity vector imaging. D, Speckle-based SR.

Tissue Doppler-based Strain

• This was the initial echocardiographic approach to measurement of myocardial deformation.
• The technique is based on comparison of the gradient of tissue velocity over the rate of change in length over a sample volume to obtain SR (ΔV/Δr). Strain is derived as the temporal integral of the spatial differential of velocity ∫(ΔV/Δr)δt.
• Steps in acquisition
• Save three cycles (some equipment uses a minimum 1-s capture).
• Optimize frame rate (>100 frames per second [fps]).
• Check tissue velocity to avoid aliasing, and modify Nyquist accordingly.
• Consider narrow sector acquisition to optimize spatial and temporal resolution, as long as whole LV is not required (e.g., stress or synchrony assessments).
• Interpretation steps
• Define start and end of systole using electrocardiography (end of T wave), aortic valve motion, color Doppler (isovolumic signal; Fig. 6-6) or LV outflow.
• Use a sample volume of 12 × 6 mm (note that data are gathered outside of this).
• Examine quality on the tissue Doppler; if there is artifact or aliasing, more complex analysis (i.e., strain) probably will not work.
• Track samples to the cardiac cycle.
• Move sample volume until a reproducible SR signal is obtained in the segment of interest and repeat in a reference segment. Measure peak systolic SR (Fig. 6-7).
• Change to strain, check timing, peak systolic strain, and postsystolic index.
• Limitations
• The fundamental limitation of tissue Doppler-based strain is directional.
• The motion that is measured should be parallel to the ultrasound beam, and generally this limits the application of the technique to apical images, although some work has been done in the anteroseptal and posterior walls from parasternal views.
• This technique is somewhat dependent on image quality, but the signal-to-noise ratio of Doppler is higher than for gray scale imaging, meaning that pictures unsuitable for tracing endocardial borders can sometimes provide accurate tissue velocity-based strain.
• The requirement for high frame rate, particularly for SR imaging, may be difficult to reconcile with a wide imaging sector.
• Integration with contrast is difficult. While the signal usually appears reliable, some studies have shown discrepancies with noncontrast images.
image

Figure 6-6 Definition of end systole from the isovolumic signal in the TVI waveform (arrow).

image

Figure 6-7 Color TVI (A), SR (B), and strain (C) from the same two cardiac cycles. Systolic deflections (arrows) are positive by tissue Doppler and negative by SR and strain.

Speckle-based Strain

• The application of 2D speckle technology on standard commercial machines is now more than 5 years old.
• How it works:
• The ultrasound image is composed of a matrix of speckles that correspond to minute tissue structures.
• A process of pattern recognition allows these to be tracked from frame to frame, allowing the assessment of strain in any direction. Speed of movement is identified on the basis of the degree of excursion from frame to frame.
• Acquisition frame rate is extremely important, and optimal rate is 60 to 80 fps. If too low, speckles risk traveling out of the image; if too fast, the excursion between frames will be too small to measure.
• The adequacy of tracking can be expressed as the correlation between images. This can be mapped, to give the reader an impression for the reliability of the data, but apart from a process of offering minor guidance about tracking by one manufacturer, this quality control measure is unexploited.
• Strain measurement on archived images can be performed using some software that will process data from all vendors. However, caution needs to be applied to using this on DICOM images or the gray scale image in the background of color tissue velocity imaging (TVI), as the temporal resolution of both are too low for speckle tracking.
• The primary data are acquired as strain. SR can be calculated as the derivative of strain over time, but some constraints arise from acquisition frame rate (Fig. 6-8).
• Acquisition
• Optimize frame rate (50–80 fps).
• Optimize image—speckle is susceptible to artifact (e.g., ribs).
• Interpretation steps
• Trace endocardial border or add guide points to permit segmentation of myocardium.
• Define timing markers (end systole).
• Check adequacy of tracking.
• Limitations
• Dependent on image quality, this may pose particular problems in the far field.
• Use after contrast left ventricular opacification is not feasible, presumably because of tracking contrast microbubbles in the myocardium.
• SR analysis is somewhat limited by temporal resolution.
image

Figure 6-8 Processing of speckle tracking to obtain strain. A high frame-rate two-dimensional image is traced at end systole (A). The wall is automatically segmented to provide velocity (B) or strain (C). Curves show the course of development of velocity or strain over time.

Speckle- versus TVI-based Strain

• The relevant features of TVI, speckle, and TVI strain are summarized in Table 6-2.
• Generally, the feasibility and angle independence of speckle strain have made this the test of choice in the assessment of myocardial deformation.
• The relative simplicity of TVI means that this should be considered when there is no requirement for site specificity, especially if timing data are sought.
• TVI-based strain is preferable if assessment of deformation is sought at high heart rate (e.g., during stress) or there is a need to measure SR.

TABLE 6-2

Related posts:

  1. Assessing Twist and Torsion
  2. Advanced Echocardiography Approaches: 3D Transesophageal Assessment of the Mitral Valve
  3. Transthoracic Three-Dimensional Echocardiography
  4. Two-Dimensional and Three-Dimensional Echocardiographic Evaluation of the Right Ventricle

Stay updated, free articles. Join our Telegram channel
Tags:
Jun 4, 2016 | Posted by in CARDIOLOGY | Comments Off on Strain and Strain Rate Imaging

Full access? Get Clinical Tree
Get Clinical Tree app for offline access