Measurement of Strain and Strain Rate by Echocardiography Ready for Prime Time?

Transcription

1 Journal of the American College of Cardiology Vol. 47, No. 7, by the American College of Cardiology Foundation ISSN /06/$32.00 Published by Elsevier Inc. doi: /j.jacc STATE-OF-THE-ART PAPER Measurement of Strain and Strain Rate by Echocardiography Ready for Prime Time? Thomas H., MD, PHD Brisbane, Australia Strain and strain rate (SR) are measures of deformation that are basic descriptors of both the nature and the function of cardiac tissue. These properties may now be measured using either Doppler or two-dimensional ultrasound techniques. Although these measurements are feasible in routine clinical echocardiography, their acquisition and analysis nonetheless presents a number of technical challenges and complexities. Echocardiographic strain and SR imaging has been applied to the assessment of resting ventricular function, the assessment of myocardial viability using low-dose dobutamine infusion, and stress testing for ischemia. Resting function assessment has been applied in both the left and the right ventricles, and may prove particularly valuable for identifying myocardial diseases and following up the treatment response. Although the evidence base is limited, SR imaging seems to be feasible and effective for the assessment of myocardial viability. The use of the technique for the detection of ischemia during stress echocardiography is technically challenging and likely to evolve further. The clinical availability of strain and SR measurement may offer a solution to the ongoing need for quantification of regional and global cardiac function. Nonetheless, these techniques are susceptible to artifact, and further technical development is necessary. (J Am Coll Cardiol 2006;47: ) 2006 by the American College of Cardiology Foundation The usual indices of global left ventricular (LV) function, such as ejection fraction and volumes, are load-dependent, and standard volumetric approaches to their measurement may be influenced by image quality, technical considerations such as off-axis imaging, and measurement error. The assessment of regional function is more difficult, remains highly subjective, and requires significant training. The echocardiographic measurement of myocardial strain ( ) offers a series of regional and global parameters that may be useful in the assessment of systolic and diastolic function. The purpose of this review is to examine the technical and clinical aspects of incorporating this measurement into daily clinical practice. TECHNICAL ASPECTS Background. Strain is a measure of tissue deformation. As the ventricle contracts, muscle shortens in the longitudinal and circumferential dimensions (a negative strain) and thickens or lengthens in the radial direction (a positive strain). The application of strain to measure deformation is constrained by a number of complexities when the parameter is measured by echocardiography. First, to quantify the From the University of Queensland Department of Medicine, Princess Alexandra Hospital, Brisbane, Australia. Supported in part by a project grant (210218) from the National Health and Medical Research Council of Australia, Canberra, Australia. The author s research group has collaborative research projects with General Electric Medical Systems. Manuscript received August 8, 2005; revised manuscript received November 21, 2005, accepted November 22, lengthening or shortening process an initial measurement of length is required (Lagrangian strain), and the same findings may not necessarily be obtained by the measurement of instantaneous strain during contraction (Eulerian or natural strain). Second, tissue deformation occurs in three planes, in addition to which shearing motion involves a number of other tensors, so our current measurement approaches are a vast simplification of the true motion of the heart. Third, the assumption that tissue is incompressible is not completely true, and for example ignores the variation in myocardial blood volume between diastole and systole. Fourth, the complexities of fiber direction cause a longitudinal shortening of 20% to 30% to generate radial shortening of 50% to 70% (1). Strain rate (SR) measures the time course of deformation, and is the primary parameter of deformation derived from tissue Doppler (see later text). Indeed, SR seems to be a correlate of rate of change in pressure (dp/dt), a parameter that is used to reflect contractility, whereas strain is an analog of regional ejection fraction (2). As would be expected with ejection fraction, increasing pre-load is associated with increasing strain at all levels of wall stress, and increasing after-load is associated with a reduction of strain. Although LV cavity size close to the normal range has a limited impact on strain, radial strain is increased and longitudinal strain is reduced in small left ventricles. In contrast, SR is thought to be less related to pre-load and after-load.

2 1314 JACC Vol. 47, No. 7, 2006 Abbreviations and Acronyms 2D two-dimensional LV left ventricle/ventricular RV right ventricle/ventricular SR strain rate SRI strain rate image/imaging Myocardial strain may be measured using a variety of echocardiographic techniques. Although M-mode techniques provide both accurate temporal and accurate spatial resolution, and may therefore be used to measure strain in a single dimension, the current era of myocardial strain measurement began with the measurement of SR from comparison of adjacent tissue velocities by Heimdal et al. (3). Subsequently, strain has been measured using speckle tracking techniques (4,5). Each of these methodologies presents its own clinical challenges. Tissue Doppler-based strain. TECHNICAL ASPECTS. The velocity of movement of myocardium can be recorded by tissue Doppler techniques and displayed as a parametric color image in which each pixel represents the velocity relative to the transducer. These data may also be expressed graphically as the velocity of the myocardium relative to time (on the x axis). These recordings have documented that a descending gradation of velocity exists from the LV base to apex, reflecting the contraction of the base toward a relatively fixed apex. Figure 1A shows the gradation of peak velocities at different locations along the LV wall. Although these velocity recordings provide information about the motion of the wall, the ability of contraction in adjacent segments to influence the velocity in any given segment limits the site-specificity of velocity data. Rather than examine the motion of a segment relative to the transducer, which is susceptible to tethering to adjacent tissue, myocardial motion may be measured relative to the Figure 1. Derivation of strain rate (SR) and strain from tissue Doppler data. A series of velocity curves (comprising isovolumic contraction [IVC], systolic [S] and diastolic [E and A] components) show a velocity gradient along a length of the wall (labeled d in the color Doppler image in A). A regression calculation between adjacent tissue velocity data points along this length generates the strain rate curve (B), which is then integrated to calculate strain (C). Timing of end-systole can be confirmed from the tissue Doppler waveform in a separate example, the aortic valve closure (AVC) is marked by a transient wave in the adjacent septum and anterior mitral leaflet (D). ES end-systolic; IVR isovolumic relaxation. Continued on next page.

3 JACC Vol. 47, No. 7, adjacent myocardium. The instantaneous gradient of velocity along a sample length may be quantified by performing a regression calculation between the velocity data from adjacent sites along the scan line, and these instantaneous data may then be combined to generate an SR curve (Fig. 1B) (6,7). Integration of this curve provides instantaneous data on deformation shortening or lengthening that represent strain (Fig. 1C). These data therefore reflect the movement of one tissue site relative to another within the sample volume, in contrast to tissue velocity data, which merely reflect movement of one site relative to the transducer. A number of experimental and clinical articles have attested to the benefits of site specificity in avoiding motion caused by tethering to adjacent segments, which is especially important when dealing with coronary artery disease (8,9). Our approach is first to examine the tissue velocity waveform, because this represents the primary data, and this approach often avoids being misled by technical problems such as excessive noise or aliasing, which preclude further analysis. The next step is to define the timing of the waveform (Fig. 1D), starting with aortic valve opening (which follows the isovolumic contraction spike on the velocity curve), aortic valve closure (marked by a transient shock wave in the septum and mitral valve), and mitral valve opening (readily detectible from gray-scale imaging in all apical views). Like tissue velocity, strain parameters are most commonly used to assess myocardial motion in a base-to-apex direction, which is sensitive to mild subendocardial damage. In contrast, the measurement for radial strain from tissue velocity data is unsuitable for clinical use. It is difficult to accommodate the optimal inter-site distance required for SR measurements (12 mm) in a ventricle of normal thickness, and the use of a shorter offset distance is associated with greater noise levels. Moreover, the requirement for the adjacent points to lie along a single regression line means that only anteroseptal and posterior segments can be analyzed with this technique, and because of the combination of right ventricular (RV) and LV myocardial structure in the septum, effectively only radial strain of the posterior wall measurements are meaningful. LIMITATIONS OF DERIVATION OF SR FROM TISSUE VE- LOCITY. The velocity-regression technique has a number of potential pitfalls (Table 1). First, the comparison of adjacent Figure 1 Continued.

4 1316 JACC Vol. 47, No. 7, 2006 Table 1. Problems and Solutions for Tissue-Velocity Based Strain Rate Imaging Problem Signal noise Underestimation Angle dependence Through-plane motion Respiratory drift TVI tissue velocity imaging. Solution Ensure clean velocity signal (avoid reverberations on two-dimensional echocardiography) Use harmonic imaging Avoid aliasing on TVI signal (use adequate pulse repetition) Track sample volume to left ventricular wall to avoid cavity signal Use high frame rate Align axis of movement with scan line Narrow sector Caution with interpretation of events late after QRS Acquire in end-expiration velocities is exquisitely sensitive to signal noise, and the quality of SR curves may vary depending on the care used in obtaining the underlying velocity data (Fig. 2). Optimizing the velocity signal should include avoidance of reverberation artifact (Fig. 3A) and ensuring adequate frame rate ( 100 frames/s). Inadequate pulse-repetition frequency leads to aliasing (Fig. 4). Improvements to the velocity signal by use of harmonic imaging as well as both temporal and spatial averaging are important in optimizing the SR signal, although this comes at the cost of reducing spatial resolution (10). The second limitation relates to the limits on spatial resolution that are imposed by imaging at high temporal resolution. If the number of Doppler interrogating beams is limited in an effort to maximize temporal resolution, spatial resolution may be compromised. This may contaminate myocardial velocity signals with adjacent LV blood pool velocities, which are an important source of noise (Fig. 3B). In turn, this will compromise the SR signal (Fig. 3C). Tracking the sample throughout the cardiac cycle is also important to ensure that the sample remains within the myocardium. Use of a narrow imaging sector although inconvenient for clinical imaging enables a limited number of Doppler beams to be focused in a small area, optimizing spatial resolution. If a narrow sector is undesirable, reduction of frame rate will allow the use of more Doppler beams across the imaging sector, effectively sacrificing temporal for spatial resolution. These limitations on lateral resolution significantly limit the ability of the technique to assess longitudinal subendocardial and subepicardial SR during standard imaging. Third, like all Doppler techniques, tissue velocity-based strain is sensitive to alignment. The application of this technique to areas where the axis of contraction changes along the scan line (e.g., the apex) means that different vectors may be involved at each site (Fig. 5), with consequent error in strain measurement (11,12). Fourth, the derivation of data along the scan line means that the velocity regression technique is unidirectional. Even when tracking is used to try to maintain the sample volume within a segment of myocardium, it needs to be remembered that the myocardium undergoes a wringing, torsional motion so that the sample will inevitably move out of the scanning field in the course of the cardiac cycle. This motion has little effect on systolic measurements, because peak SR occurs early in systole, but it may become important in the measurement of diastolic phenomena. These considerations of through-plane motion may be particularly important when the myocardial function is non-uniform, as for example, with an ischemic cardiomyopathy. Finally, angle changes during the cardiac cycle and with respiratory movement may contribute to drifting of the strain curve. These technical challenges of tissue velocitybased SR measurements can be avoided by careful acquisition (Table 1). VALIDATION. Despite these limitations, it is important to acknowledge that this technique has been extensively validated, initially with sonomicrometry (11). Subsequent studies have confirmed correlation with magnetic resonance imaging (13). Echocardiography-based measurement of strain. RA- TIONALE. A Doppler-independent technique for strain measurement would have attractions with respect to signal noise, angle dependency, and the ability to monitor Figure 2. Variations in strain rate (SR) signal quality. A good-quality curve has well-defined components and limited signal noise (A). Increasing degrees of signal noise compromise the definition of peak SR (initially influencing timing parameters, (B) and if sufficiently severe may preclude even the measurement of amplitude parameters (C).

5 JACC Vol. 47, No. 7, Figure 3. Pitfalls of tissue Doppler-derived strain rate. (A) Reverberation (here related to rib artifact marked by an arrow and shown on the yellow curve) compromises the strain rate signal, contrasted with an adjacent normal strain rate signal (blue curve). (B) The importance of avoiding blood-pool activity, in which a noisy strain rate curve (yellow) is compared with a smaller sample size, tracked to myocardial movement (blue). (C) The limited spatial resolution of tissue Doppler, in which the blue curve (sample volume outside of the cardiac contour), although noisy, is comparable with the yellow curve that is appropriately tracked to the wall.

6 1318 JACC Vol. 47, No. 7, 2006 Figure 4. The impact of aliased peak stress tissue velocity data on strain rate calculation. (A) The presence of aliased tissue Doppler data is identified by the golden coloration of the inferior wall on the color tissue Doppler image (lower left), which in turn produces a mottled appearance of the color strain rate image (upper left) and a meaningless strain rate curve (note the absence of negative systolic deflection [arrows]). (B) The tissue velocity data have been gathered with a higher aliasing velocity note the different color map appearances and typical waveform. strain in two dimensions rather than one dimension. Various echocardiographic techniques have been used, including comparison of adjacent radiofrequency signals (4), and more recently, block-matching and speckle tracking techniques (14,15). These speckles are ultrasound reflectors within tissue, are highly reproducible, and essentially behave like magnetic resonance tags (Fig. 6). Shortening may be calculated by comparison of these speckles from frame to frame, although attention to technical detail is important, because comparisons at high frame rates are associated with high levels of noise, and comparisons at low frame rates risk loss of correlation because of excessive displacement of the speckles. COMPARISON WITH TISSUE VELOCITY METHODOLOGY. Because the assessment incorporates baseline length, twodimensional (2D) strain (in contrast to the velocity regression approach) is able to measure Lagrangian strain. The approach also has the added attraction of offering a feasible approach to radial measurements, which may be a more accurate measure of wall thickening than M-mode echocardiography, in which a proportion of apparent thickening is thought to reflect joining of the trabeculae. Finally, the technique offers a completely new approach to the assessment of torsional motion, derived from circumferential strain at different levels in the heart (16). However, differences in frame rate and smoothing lead to the availability of less detail in the SR and strain curves, with potential difficulties in the measurement of timing parameters. Nonetheless, magnitude parameters seem to be analogous with the 2D strain and velocity regression approaches (14),

7 JACC Vol. 47, No. 7, Figure 5. Impact of angulation on strain rate imaging. Interrogation parallel with the wall (mid-septum, shown in blue) identifies long-axis shortening, and at right angles to the wall (apex, shown in red) identifies short-axis thickening. However, an intermediate angle (apical septum, shown in yellow) causes underestimation a mixture of vectors at 45% produces a net absence of recordable strain. Scan planes are shown as continuous lines, longitudinal and radial contraction vectors as broken lines. and the initial experience with 2D strain methodologies suggests that they are robust, although the evidence base is limited and additional clinical assessment is required. Optimal parameter. The availability of a number of display techniques and a host of potentially measurable parameters has led to a bewildering level of choice for the novice user. Generally, these parameters can be separated into those relating to the timing and magnitude of contraction (Table 2). No parameter is suitable for all applications. TIMING PARAMETERS. The definition of these timing parameters is essential to distinguish peak systolic SR and strain from peak SR and strain these will be different in the presence of post-systolic thickening. The degree of this thickening is expressed as the post-systolic index (i.e., post-systolic increment divided by systolic strain) (17,18). Post-systolic motion is reported in 30% of myocardial segments in normal subjects, but can be identified as pathologic if there is a concomitant reduction of systolic strain, especially if the post-systolic thickening is marked (e.g., index 25% to 35%). Although pathologic postsystolic thickening usually reflects ischemic or viable myocardium, this entity may be seen in other myocardial pathologies such as LV hypertrophy, and may occur as a passive phenomenon in dyskinetic segments (19). The other timing parameters are the time until the onset of systole and the time to relaxation. Both are prolonged in Figure 6. Two-dimensional (2D) strain is based on comparison of the image texture (i.e., pattern of individual speckle elements) from frame to frame. The distortion of this pattern permits assessment of strain in the axis of movement rather than the axis of the ultrasound beam.

8 1320 JACC Vol. 47, No. 7, 2006 Table 2. Parameters Obtained From Strain Rate Imaging Type of Parameter Specific Measurement Comment Timing Time to onset of systole and time to relaxation Problems with reproducibility in clinical settings, may be less suited to 2D strain due to lower temporal resolution. Magnitude (longitudinal or radial) Peak systolic strain rate Peak diastolic strain rate Peak systolic strain Susceptible to angulation issues with TVI-based strain, probably more robust with 2D strain but frame rate may be a limitation. Through-plane motion may limit site specificity. Corresponds with regional ejection fraction and may be less suited to stress echocardiography. Magnitude timing End-systolic strain Dependent on accurate defining of end-systole, lacks information about rate of contraction. Post-systolic thickening Not a specific marker of ischemia. 2D two-dimensional; TVI tissue velocity imaging. ischemic myocardium, but although delayed contraction is a pathophysiological hallmark of ischemia, and despite some favorable clinical results (20), the clinical application of timing parameters is limited by measurement variations. It seems more feasible to obtain these results using tissue velocity signals, which, although not site-specific, are less prone to artifact. For example, the prediction of reverse remodeling after cardiac resynchronization therapy seems to be more consistently identified using tissue velocity rather than SR techniques (21). MAGNITUDE PARAMETERS. Normal ranges of SR and strain have been described (22). Normal resting values for longitudinal SR vary between 1.0/s and 1.4/s, with the standard deviation in most locations ranging from 0.5/s to 0.6/s. Normal longitudinal systolic strain in most segments varies from 15% to 25%, with normal radial strains ranging from 50% to 70%, and standard deviations of 5% to 7% (22). Although reproducibility data have been published, there has been little attention to test/retest variation, which is important if the technique is to be used in serial follow-up. Normal ranges for magnitude parameters are influenced by increasing age, pre-load (strain increases as LV size increases), and after-load (strain decreases with increasing blood pressure). Strain rate seems to be less dependent on loading. Regional variations pose an even greater problem in addition to ischemia, these may be caused by curvature or by non-uniform fiber direction and differences in angulation (22). CLINICAL APPLICATIONS LV function. Maximal elastance, based on creating pressure-volume loops at various levels of pressure and volume through alteration of pre-load and after-load, is the gold standard for the global assessment of LV function, albeit of limited clinical feasibility. Comparison of SR imaging (SRI) with elastance has shown very high correlation with peak and mean SRs, rather than strain or tissue velocity (23). However, variations of magnitude measurements caused by signal noise and the influence of different hemodynamic settings both pose a significant challenge to finding a normal range to permit comparisons between individuals (22). These problems are compounded by interand intra-observer variability in measurements (r values range from 0.7 to 0.8, with mean differences of 0.10 for SR and 1% for strain). Initial data with the use of 2D strain to assess global LV function suggest that this technique may be more feasible (5). MYOCARDIAL DISEASE. Despite these limitations, the sensitivity of SR has made it a very effective tool in the evaluation of subclinical heart disease (Table 3) (24 31). In particular, the technique has been valuable in the detection of myocardial involvement in non-cardiac diseases such as amyloidosis, diabetic heart disease, and Friedrich ataxia (24), and in the distinction of hypertrophy caused by hypertension and cardiomyopathy (27,32). Strain has been very effective in settings in which the parameter has been compared within the same individual, such as examining the treatment response to hypertensive heart disease (33), diabetes (34) and Fabry disease (35). Nonetheless, in the setting of global disease, in which the site specificity of SR imaging is not required, it is unclear whether there is a specific advantage in using SR in preference to tissue velocity, and in our own experience of assessing subclinical Table 3. Use of Strain ( ) and Strain Rate (SR) for Detection of Subclinical Disease of the Myocardium Disease Evidence Amyloidosis Systolic /SR (not S= or E=) identified subclinical disease (24) Friedrich ataxia Reduction of LV (not RV), systolic /SR, and diastolic SR proportionate to LVH (25) LV hypertrophy Systolic /SR reduced in LVH before abnormal filling (26), systolic distinguished HCM and HT-LVH (27) Tetralogy RV-S and /SR reduced (28), IVA (not S= or SR) proportionate to PR severity (29) Senning Reduction of systolic /SR in the systemic RV (30) Valvular disease Subclinical LV dysfunction in asymptomatic MR (31) E= early diastolic tissue velocity; HCM hypertrophic cardiomyopathy; HT-LVH hypertensive left ventricular hypertrophy; IVA isovolumic acceleration; LV left ventricle/ventricular; LVH left ventricular hypertrophy; MR mitral regurgitation; PR pulmonic regurgitation; RV right ventricle/ventricular; S= tissue systolic velocity; SRI strain rate image/imaging.

9 JACC Vol. 47, No. 7, Table 4. Deformation at Rest and With Dobutamine for Various Ischemic Entities Strain Rate Strain Post-Systolic Index Rest Dobutamine Rest Dobutamine Rest Dobutamine Ischemic Reduced Worsens Reduced Worsens Increased Increases more Stunned Reduced Improves Reduced Improves Increased Decreases Non-transmural infarction Reduced Biphasic Reduced No change Increased Increases more Transmural infarction Very reduced No change Very reduced No change Increased No change diabetic heart disease (36), tissue E velocity seems at least as sensitive as SRI markers. VALVULAR HEART DISEASE. Measurement of myocardial function may be important in understanding the physiological impact of valvular heart disease. A recent report on percutaneous heart valve replacement showed dramatic improvements of SR and strain (37). Subclinical myocardial dysfunction may be identified as a potential guide to the timing of surgical intervention in regurgitant valve lesions (31). DIASTOLIC DYSFUNCTION. The current evaluation of diastolic impairment is on the basis of altered transmitral flow, which is load-dependent. A direct myocardial parameter for identifying abnormal relaxation as well as its response to therapy could be of significant value. However, LV filling is influenced not only by the magnitude of diastolic SR but also by the base-to-apex propagation velocity of the relaxation wave (38,39), which is load-dependent. Moreover, the role of cross-plane movement, which may move the original sample volume outside of the imaging field, is likely to be more important for diastolic than systolic measurements. Therefore, although tissue velocities have been useful in separating normal from pseudonormal LV filling, the incremental value of strain and strain imaging to diastolic evaluation remains unclear. Right ventricular function. Measurement of RV strain and SR, although currently feasible (1) and certainly of Figure 7. Quantification of the biphasic response using strain rate imaging. This dobutamine echo was performed 3 months after anterior infarction. Resting images are in the upper left with 5 g intheupper right, 10 g inthelower left, and 40 g inthelower right. The two-dimensional images (A) show a resting wall motion abnormality in the apical septal and lateral walls, both of which seem to improve at low dose and deteriorate at peak dose. For an accompanying video, please see the Appendix. Continued on next page.

10 1322 JACC Vol. 47, No. 7, 2006 potential interest in the evaluation of congenital heart disease, remain challenging. The tissue Doppler approach to radial strain measurement is difficult because the RV wall is too thin to permit an adequate regression distance, and the place of 2D strain is undefined in this respect. Strain assessment of the septum is complicated because of RV and LV components, so the long-axis assessment of RV function is best performed in the free wall, using apical imaging. Strain measurements are higher than in the LV, and increase from base to apex. Strain techniques have been used to identify the myocardial sequelae of congenital heart disease, specifically, dysfunction of the systemic right ventricle in patients treated with the Senning procedure for transposition, or RV dysfunction caused by pulmonary regurgitation after surgery for Fallot tetralogy (28 30). Such findings may be used to confirm the presence of otherwise ambiguous findings, including ventricular non-compaction and arrhythmogenic RV dysplasia. Low-dose stress responses. The responses of strain and SR to stress have been extensively studied in animal models. In normal myocardium, increasing doses of dobutamine are associated with increasing SR throughout the study, but in contrast, myocardial strain initially increases and then decreases as heart rate increases (40). These changes have been used to argue that SR is the preferred parameter for the assessment of myocardial function during stress, although they do not account for the greater technical challenge of measuring SR during stress, nor the degree of differences that occur in strain measurements. Table 4 summarizes the short-axis strain and SR responses of different myocardial entities. At rest, stunned and Figure 7 Continued. Strain profiles (B) of the apical lateral segment show lengthening at rest, shortening at low dose, and lengthening at peak stress, and a similar pattern of augmentation and deterioration is apparent on the strain rate curves. There are no changes in the apical septal segment, consistent with infarction. Gadolinium-contrast magnetic resonance confirmed the presence of scar in the apex.

11 JACC Vol. 47, No. 7, Figure 8. Quantification of ischemic changes using strain rate imaging. In these views of a standard dobutamine echocardiograph, resting images are shown in the upper left, 10 g intheupper right, 30 g inthelower left, and 40 g inthelower right. End-systolic frames of the apical two-chamber (A) and four-chamber (B) views are shown, inferior and septal walls are outlined, and stress-induced wall motion abnormalities in the basal inferior (A) and septal (B) segments are marked by arrows in the pre-peak and peak dose images. For an accompanying video, please see the Appendix. Continued on next page. acutely ischemic myocardium and non-transmural infarction are associated with reduction of strain and SR, together with the presence of post-systolic thickening. Transmural infarction is associated with lower strain and SR and less post-systolic thickening then these other entities (41). Low-dose dobutamine increases the strain and SR and reduces the post-systolic thickening in stunned myocardium, but non-transmural infarcts show only a transient increase of SR, no change of strain, and increasing postsystolic thickening. As might be expected, acutely ischemic tissue deteriorates and transmural infarction remains unchanged. Because both stunned myocardium and transmural infarction may show marked reduction of systolic thickening, other investigators have made the distinction on the

12 1324 JACC Vol. 47, No. 7, 2006 Figure 8 Continued. Strain and strain rate profiles in the basal inferior (C) and septal (D) segments show reduction of basal strain ( 10%) and strain rate ( 1/s) at peak stress, with no changes in the mid/apical segments. Continued on next page.

13 JACC Vol. 47, No. 7, Figure 8 Continued. Coronary angiography (E) shows significant right coronary disease. basis of differences of diastolic strain and SR. Limited information has been obtained using apical imaging, but strain studies in animal models have suggested differences in subendocardial and mid-myocardial contraction (42), although it is unclear whether the lateral resolution of the SRI technique is able to provide this information at clinical imaging depths. The clinical application of SRI to the assessment of viability presents some important differences in comparison with the animal models. First, SR measurement in the clinical setting in the thick chest wall may yield suboptimal images and is technically more challenging. Second, the predominant use of short-axis imaging in the animal model is difficult to apply clinically for the reasons discussed above. Nonetheless, the improvement and deterioration of viable myocardium showing a biphasic response may be quantified using SR imaging (Fig. 7). Moreover, validation of SRI for the assessment of myocardial viability has been reported in two studies. Hoffmann et al. (43) studied 37 patients with ischemic LV dysfunction who underwent low-dose dobutamine stress echocardiography and positron emission tomography. Viability was defined on the basis of perfusion metabolism mismatch, and SRI was found to be more accurate than tissue velocity imaging in the prediction of viability (area under the receiver operating characteristic curve 0.89 vs. 0.63). An optimal cutoff of SR increment of 0.23 predicted viability with a sensitivity of 83% and specificity of 84%. A follow-up study showed viable segments to augment diastolic SR waves with dobutamine (44). Hanekom et al. (45) used the recovery of segmental function after revascularization to define myocardial viability. Similar areas under the receiver operating characteristic curves for strain and SR were obtained as in the study of Hoffmann et al. (43), with a similar increment of 0.25/s of SR, and low-dose SR of 1/s being the optimal cutoff for the prediction of functional recovery. Although Figure 9. Use of two-dimensional (2D) strain to quantify changes in wall thickening in ischemic territories. The baseline images showed apical septal hypokinesis, manifest as reduced amplitude of the strain-rate curve (green, marked with short arrow), but this worsened to display systolic lengthening and post-systolic shortening with stress (long arrow). The patient had critical disease of the left anterior descending artery at coronary angiography.

14 1326 JACC Vol. 47, No. 7, 2006 these parameters were not superior to wall motional assessment, they were incremental to its use. Thus, these two limited clinical studies suggest that SR imaging is feasible during low-dose dobutamine, and that an increment in SR, as shown in the animal studies, is predictive of viability. Assessment of myocardial ischemia. The requirement for specific training and the ongoing, albeit reduced, discordance between expert readers both make the application of SR imaging to stress echocardiography a very desirable goal. Tissue velocity imaging has already been applied to stress echocardiography, with favorable improvement in concordance between observers and enhancement of the accuracy of novice readers (46). However, although the tissue velocity technique may be able to distinguish normal from abnormal, its lack of site specificity makes it unattractive for the assessment of coronary artery disease, and passive motion bestowed by tethering of adjacent normal segments may compromise the sensitivity of this method. Although SRI may have the benefit of better specificity for location, it is technically challenging and for this reason has not been considered to be feasible with exercise echocardiography (47). Experimental studies have suggested that the measurement of SR, time to relaxation, and degree of post-systolic thickening are suitable markers for the detection of ischemia. Human studies obtained during angioplasty (18) proposed an optimal longitudinal SR of 0.8/s (sensitivity 75%, specificity 63%) and a peak systolic strain of 10% (sensitivity 86%, specificity 83%) as suitable cutoffs for detection of ischemia. Although both were superior to tissue velocity measurement, the optimal longitudinal parameter was a post-systolic strain index of 0.25 (sensitivity 95%, specificity 89%). Unfortunately, the intensity of supplyside ischemia during angioplasty exceeds the intensity of ischemia during stress testing, and the non-ischemic segments are in a very different milieu. The only clinical article to examine SR imaging during dobutamine stress echocardiography (40) showed strain cutoffs to be insensitive, whereas peak systolic SR was the optimal magnitude parameter, and post-systolic thickening was the optimal measurement for the detection of ischemia. A postsystolic shortening cutoff of 35% gave a sensitivity of 82% and specificity of 85% for the detection of ischemia, and the visual impression of both timing and the magnitude of contraction using SRI in anatomic M-mode exceeded the accuracy wall motion assessment. However, in our experience, changes in strain and SR magnitude caused by ischemia may be more striking than timing changes (Fig. 8). The use of SRI at peak-dose dobutamine may be associated with compromise of signal quality because of artifact, and 2D strain may be more feasible for this purpose (Fig. 9). CONCLUSIONS Over its five-year history, SRI has provided a valuable physiological tool for understanding myocardial mechanics. Unlike its parent methodology, tissue Doppler imaging, which has found a niche in the assessment of diastolic dysfunction and measurement of LV synchrony, the place of SRI in standard clinical practice remains incompletely defined. The most immediate clinical applications relate to myocardial viability and the identification of subclinical LV dysfunction, with the application of standard stress echocardiography and quantification of resting function being more remote goals. Barriers to the clinical uptake of this technique include the requirement for significant understanding of complex methodology, technical challenges of acquisition and analysis, and lack of consensus regarding the superiority of any one among a number of potential measurements for different applications. Ongoing research will be required to clarify the true value of this interesting and promising modality as a routine clinical tool. Reprint requests and correspondence: Dr. Thomas H., University of Queensland Department of Medicine, Princess Alexandra Hospital, Brisbane, Qld 4102, Australia. tmarwick@soms.uq.edu.au. REFERENCES 1. Kowalski M, Kukulski T, Jamal F, et al. Can natural strain and strain rate quantify regional myocardial deformation? A study in healthy subjects. Ultrasound Med Biol 2001;27: Weidemann F, Jamal F, Sutherland GR, et al. Myocardial function defined by strain rate and strain during alterations in inotropic states and heart rate. Am J Physiol Heart Circ Physiol 2002;283:H Heimdal A, Stoylen A, Torp H, Skjaerpe T. Real-time strain rate imaging of the left ventricle by ultrasound. J Am Soc Echocardiogr 1998;11: D hooge J, Konofagou E, Jamal F, et al. Two-dimensional ultrasonic strain rate measurement of the human heart in vivo. IEEE Trans Ultrason Ferroelectr Freq Control 2002;49: Reisner SA, Lysyansky P, Agmon Y, Mutlak D, Lessick J, Friedman Z. Global longitudinal strain: a novel index of left ventricular systolic function. J Am Soc Echocardiogr 2004;17: D hooge J, Heimdal A, Jamal F, et al. Regional strain and strain rate measurements by cardiac ultrasound: principles, implementation and limitations. Eur J Echocardiogr 2000;1: Sutherland GR, Di Salvo G, Claus P, D hooge J, Bijnens B. Strain and strain rate imaging: a new clinical approach to quantifying regional myocardial function. J Am Soc Echocardiogr 2004;17: Armstrong G, Pasquet A, Fukamachi K, Cardon L, Olstad B, T. Use of peak systolic strain as an index of regional left ventricular function: comparison with tissue Doppler velocity during dobutamine stress and myocardial ischemia. J Am Soc Echocardiogr 2000;13: Abraham TP, Nishimura RA, Holmes DR Jr., Belohlavek M, Seward JB. Strain rate imaging for assessment of regional myocardial function: results from a clinical model of septal ablation. Circulation 2002;105: D hooge J, Bijnens B, Thoen J, Van de Werf F, Sutherland GR, Suetens P. Echocardiographic strain and strain-rate imaging: a new tool to study regional myocardial function. IEEE Trans Med Imaging 2002;21: Urheim S, Edvardsen T, Torp H, Angelsen B, Smiseth OA. Myocardial strain by Doppler echocardiography. Validation of a new method to quantify regional myocardial function. Circulation 2000; 102:

15 JACC Vol. 47, No. 7, Storaa C, Aberg P, Lind B, Brodin LA. Effect of angular error on tissue Doppler velocities and strain. Echocardiography 2003;20: Edvardsen T, Gerber BL, Garot J, Bluemke DA, Lima JA, Smiseth OA. Quantitative assessment of intrinsic regional myocardial deformation by Doppler strain rate echocardiography in humans: validation against three-dimensional tagged magnetic resonance imaging. Circulation 2002;106: Leitman M, Lysyansky P, Sidenko S, et al. Two-dimensional strain a novel software for real-time quantitative echocardiographic assessment of myocardial function. J Am Soc Echocardiogr 2004;17: Kaluzynski K, Chen X, Emelianov SY, Skovoroda AR, O Donnell M. Strain rate imaging using two-dimensional speckle tracking. IEEE Trans Ultrason Ferroelectr Freq Control 2001;48: Notomi Y, Lysyansky P, Setser RM, et al. Measurement of ventricular torsion by two-dimensional ultrasound speckle tracking imaging. J Am Coll Cardiol 2005;45: Voigt JU, Lindenmeier G, Exner B, et al. Incidence and characteristics of segmental postsystolic longitudinal shortening in normal, acutely ischemic, and scarred myocardium. J Am Soc Echocardiogr 2003;16: Kukulski T, Jamal F, Herbots L, et al. Identification of acutely ischemic myocardium using ultrasonic strain measurements. A clinical study in patients undergoing coronary angioplasty. J Am Coll Cardiol 2003;41: Skulstad H, Edvardsen T, Urheim S, et al. Postsystolic shortening in ischemic myocardium: active contraction or passive recoil? Circulation 2002;106: Abraham TP, Belohlavek M, Thomson HL, et al. Time to onset of regional relaxation: feasibility, variability and utility of a novel index of regional myocardial function by strain rate imaging. J Am Coll Cardiol 2002;39: Yu CM, Fung JW, Zhang Q, et al. Tissue Doppler imaging is superior to strain rate imaging and postsystolic shortening on the prediction of reverse remodeling in both ischemic and nonischemic heart failure after cardiac resynchronization therapy. Circulation 2004;110: Sun JP, Popovic ZB, Greenberg NL, et al. Noninvasive quantification of regional myocardial function using Doppler-derived velocity, displacement, strain rate, and strain in healthy volunteers: effects of aging. J Am Soc Echocardiogr 2004;17: Greenberg NL, Firstenberg MS, Castro PL, et al. Doppler-derived myocardial systolic strain rate is a strong index of left ventricular contractility. Circulation 2002;105: Koyama J, Ray-Sequin PA, Falk RH. Longitudinal myocardial function assessed by tissue velocity, strain, and strain rate tissue Doppler echocardiography in patients with AL (primary) cardiac amyloidosis. Circulation 2003;107: Weidemann F, Eyskens B, Mertens L, et al. Quantification of regional right and left ventricular function by ultrasonic strain rate and strain indexes in Friedreich s ataxia. Am J Cardiol 2003;91: Yuda S, Short L, Leano R, TH. Myocardial abnormalities in hypertensive patients with normal and abnormal left ventricular filling: a study of ultrasound tissue characterization and strain. Clin Sci (Lond) 2002;103: Kato TS, Noda A, Izawa H, et al. Discrimination of nonobstructive hypertrophic cardiomyopathy from hypertensive left ventricular hypertrophy on the basis of strain rate imaging by tissue Doppler ultrasonography. Circulation 2004;110: Weidemann F, Eyskens B, Mertens L, et al. Quantification of regional right and left ventricular function by ultrasonic strain rate and strain indexes after surgical repair of tetralogy of Fallot. Am J Cardiol 2002;90: Frigiola A, Redington AN, Cullen S, Vogel M. Pulmonary regurgitation is an important determinant of right ventricular contractile dysfunction in patients with surgically repaired tetralogy of Fallot. Circulation 2004;110:II Eyskens B, Weidemann F, Kowalski M, et al. Regional right and left ventricular function after the Senning operation: an ultrasonic study of strain rate and strain. Cardiol Young 2004;14: Lee R, Hanekom L, TH, Leano R, Wahi S. Prediction of subclinical left ventricular dysfunction with strain rate imaging in patients with asymptomatic severe mitral regurgitation. Am J Cardiol 2004;94: Andersen NH, Poulsen SH, Eiskjaer H, Poulsen PL, Mogensen CE. Decreased left ventricular longitudinal contraction in normotensive and normoalbuminuric patients with type II diabetes mellitus: a Doppler tissue tracking and strain rate echocardiography study. Clin Sci (Lond) 2003;105: Mottram PM, Haluska B, Leano R, Cowley D, Stowasser M, TH. Effect of aldosterone antagonism on myocardial dysfunction in hypertensive patients with diastolic heart failure. Circulation 2004;110: Voigt JU, vonbibra H, Daniel WG. New techniques for the quantification of myocardial function: acoustic quantification, color kinesis, tissue Doppler and strain rate imaging. Z Kardiol 2000;89 Suppl 1: Weidemann F, Breunig F, Beer M, et al. Improvement of cardiac function during enzyme replacement therapy in patients with Fabry disease: a prospective strain rate imaging study. Circulation 2003;108: Fang ZY, Schull-Meade R, Downey M, Prins J, TH. Determinants of subclinical diabetic heart disease. Diabetologia 2005; 48: Bauer F, Eltchaninoff H, Tron C, et al. Acute improvement in global and regional left ventricular systolic function after percutaneous heart valve implantation in patients with symptomatic aortic stenosis. Circulation 2004;110: Voigt JU, Lindenmeier G, Werner D, et al. Strain rate imaging for the assessment of preload-dependent changes in regional left ventricular diastolic longitudinal function. J Am Soc Echocardiogr 2002;15: Stoylen A, Skjelvan G, Skjaerpe T. Flow propagation velocity is not a simple index of diastolic function in early filling. A comparative study of early diastolic strain rate and strain rate propagation, flow and flow propagation in normal and reduced diastolic function. Cardiovasc Ultrasound 2003;1: Voigt JU, Exner B, Schmiedehausen K, et al. Strain-rate imaging during dobutamine stress echocardiography provides objective evidence of inducible ischemia. Circulation 2003;107: Weidemann F, Dommke C, Bijnens B, et al. Defining the transmurality of a chronic myocardial infarction by ultrasonic strain-rate imaging: implications for identifying intramural viability: an experimental study. Circulation 2003;107: Hashimoto I, Li X, Hejmadi BA, Jones M, Zetts AD, Sahn DJ. Myocardial strain rate is a superior method for evaluation of left ventricular subendocardial function compared with tissue Doppler imaging. J Am Coll Cardiol 2003;42: Hoffmann R, Altiok E, Nowak B, et al. Strain rate measurement by Doppler echocardiography allows improved assessment of myocardial viability inpatients with depressed left ventricular function. J Am Coll Cardiol 2002;39: Hoffmann R, Altiok E, Nowak B, et al. Strain rate analysis allows detection of differences in diastolic function between viable and nonviable myocardial segments. J Am Soc Echocardiogr 2005;18: Hanekom L, Jenkins C, Short L, TH. Accuracy of strain rate techniques for identification of viability at dobutamine stress echo: a follow-up study after revascularization. Circulation 2005;112: Fathi R, Cain P, Nakatani S, Yu HC, TH. Effect of tissue Doppler on the accuracy of novice and expert interpreters of dobutamine echocardiography. Am J Cardiol 2001;88: Davidavicius G, Kowalski M, Williams RI, et al. Can regional strain and strain rate measurement be performed during both dobutamine and exercise echocardiography, and do regional deformation responses differ with different forms of stress testing? J Am Soc Echocardiogr 2003;16: APPENDIX For accompanying videos to Figures 7A, 8A, and 8B, please see the online version of this manuscript.