Physiological basis in the assessment of myocardial mechanics using speckle-tracking echocardiography 2D. Part II.

In this paper, the authors attempt to concisely present the anatomical and pathophysiological bases as well as the principles for echocardiographic evaluation of mechanical aspects of cardiac function based on speckle tracking method. This technique uses a phenomenon involving the formation of characteristic image units, referred to as speckles or acoustic markers, which are stable during cardiac cycle, on a two-dimensional echocardiographic picture. Changes in the position of these speckles throughout the cardiac cycle, which are monitored and analyzed semi-automatically by a computer system, reflect deformation of both, cardiac ventricle as a whole as well as its individual anatomical segments. The values of strain and the strain rate, as well as the range and velocity of the movement of these markers, which are in close relationship with multiple hemodynamic parameters, can be visualized as various types of charts - linear, two- and three-dimensional - as well as numerical values, enabling deeper insight into the mechanical and hemodynamic aspects of cardiac function in health and disease. The use of information obtained based on speckle tracking echocardiography allows to understand previously unclear mechanisms of physiological and pathophysiological processes. The first part of the study discusses the formation of a two-dimensional ultrasound image and the speckles, as well as the technical aspects of tracking their movement. The second part presents in more detail the methodology of speckle-tracking echocardiography, the characteristic abnormalities of cardiac mechanics presenting in different clinical entities, and the limitations related to given clinical and technical issues.

Carrying out the procedure

Two-dimensional speckle tracking technique allows for an assessment of changes in the position of the individual speckles (acoustic markers) on the surface of an imaging sector, therefore an analysis of several sections perpendicular to each other is necessary to obtain complete information on the three-dimensional mechanics of a given heart chamber. Analysis of left ventricular function requires optimal twodimensional images that allow for an assessment of the individual motion components in three basic apical views: fourchamber, two-chamber and a view of LV in the long axis, as well as transverse views: visualizing LV at the level of mitral valve cusps, at the level of papillary muscles and in the apical region(. The highest possible frame rate (most authors recommend values above 60 frames/sec, and in case of tachycardia – higher values, which in turn requires the use of the narrowest possible angle of image sector – only slightly wider than the structure examined() should be used for the most accurate and smooth tracking of the movement of individual speckles. Optimal visualization of epicardium and epicardium as well as elimination of all image artifacts that could imitate the actual speckles and impair the analysis of their movement, seems necessary(. Finally, maximum stability of the position of the examined structure inside the image should be achieved to avoid the ‘loss’ of its fragments due to heart dislocation relative to chest walls. A region of interest is selected on the registered image using a special cursor, i.e. the borders of endocardium and epicardium, ventricular wall and the septum are delineated as accurately as possible, excluding the pericardium from the examined region. Once the appropriateness of the selected boundaries of the examined area is approved, the area is automatically divided by the system into six segments corresponding to left ventricular segments, and an analysis is performed. The obtained results are presented in the form of line graphs for the movement parameters of the individual segments (graph for each segment is presented in a different color) and larger regions in the course of the cardiac cycle, two-dimensional color maps of movement overlaid on the basic image, two-dimensional diagrams showing the movement of segments on the background of the whole ventricle and, finally, multiple numerical values of the individual evaluated parameters( (Fig. 1).

Fig. 1

A typical example of graphical presentation of basic data obtained in the analysis of longitudinal strain. Four images are presented in the figure: 1) ROI divided into segments and laid on the two-dimensional image of the left ventricle in the fourchamber apical view; 2) peak systolic strain values on the image of each segment; 3) color-coded line graphs for strain of each segment – the colors correspond to the segments in the two-dimensional image – and a white dotted graph for averaged global strain of the left ventricle; 4) a two-dimensional ribbon graph illustrating the strain of all segments during the cardiac cycle. The ribbon graph illustrates the course of strain in the following segments: basal – septal (at the top of the graph) – middle segments – apical (in the center) – up to the lateral segment located at the bottom of the graph. Ribbons corresponding to the individual segments are color-coded and separated by horizontal lines. The points of peak strain for each segment as well as the global strain are highlighted using small, white squares

It is also possible to calculate the ejection fraction and the stroke volume based on the combined analysis of the individual sections. Segments whose movement could not be properly traced are excluded from the analysis, which is displayed by the system on the appropriate charts. This al-lows for a manual adjustment of borders of the region of interest or rejection of inadequate sections and a reaquisition of the same view in the case of failure to achieve an adequate quality of results. Then it is possible to continue further processing of the obtained results, allowing for an assessment of the spatial differences in the kinetics of the individual myocardial regions, which is used in the diagnosis of segmental systolic dysfunction or systolic asynchrony in various diseases unrelated to focal damage, e.g. in patients after surgeries of complex congenital heart defects. The presence of characteristic changes in the curves of movement dynamics in various disease states, such as ischemic heart disease, pulmonary and systemic arterial hypertension, cardiomyopathies, etc. allows for their diagnosis, staging and monitoring of the course of treatment based on repeated STE testing.

STE – assessable basic parameters of cardiac function

Displacement (D) – a parameter specifying the distance travelled by the analyzed cardiac fragment (acoustic marker) between two subsequent image frames as well throughout the whole period of observation. The value of displacement is expressed in length units (cm)( (Fig. 2).

Fig. 2

Displacement (D) – the parameter specifies the distance travelled by the analyzed cardiac fragment (acoustic marker) between two subsequent image frames as well throughout the whole period of observation. The value of displacement is expressed in length units (cm). Radial strain is analyzed relative to the midpoint of the left ventricle defined by the intersection of the main axes in LV cross-section. The centripetal motion is red-coded in two-dimensional maps and identified as a positive deflection in line graphs; the centrifugal movement is blue-coded and identified as negative deflection.

Velocity, V – expresses the speed of the speckle moving: actual or average – in cm/s( (Fig. 3).

Fig. 3

Velocity (V) – expresses the speed of the speckle moving: actual or average – in cm/s. For longitudinal strain, the analysis is performed in the apex-base direction. The movement towards the apex is red-coded on two-dimensional maps and identified as positive deflection on line graphs; the movement towards the base is blue-coded and identified as negative deflection. As can be seen on the figure, an assessment of a single wall (septum in this case) or other architectural cardiac fragment, is also possible. Differences in the microcardial velocities in the individual segments are noticeable, with the highest velocities registered in the basal segment and the lowest – in the apical segment. However, maximum values are achieved at very similar time points

Strain (S) – describes deformation as the degree of change in the length of myocardial segment. It is a dimensionless number expressed in percentages or as a fraction of its original value. It may be positive or negative, e.g. is a 10-cm thread is stretched to a length of 12 cm, the strain is +0.2 or +20%; if the thread shrinks to 8 cm, the strain is −0.2 or −20%( (Fig. 4).

Fig. 4

Strain (S) – describes deformation as the degree of change in the length of myocardial segment. It is a dimensionless number expressed in percentages or as a fraction of its original value. Line graphs for the curse of longitudinal strain in the individual segments as well as averaged global strain and two-dimensional ribbon graph of strain may be seen in the figure

Strain rate (SR) – defines strain dynamics, i.e. change in the length per time unit, and is expressed in s−1( (Fig. 5).

Fig. 5

Strain rate (SR) – defines strain dynamics, i.e. change in the length per time unit, and is expressed in s−1. In addition to the graphs illustrating SR changes during cardiac cycle, a table containing SR at time points of its highest values, i.e. peak systolic velocity (peak S), the phase of rapid ventricular filling (peak E) and peak atrial systolic velocity (peak A), is shown for each of the analyzed segments

The movement of cardiac structures takes place in the space, therefore both the displacement and the velocity of this displacement are vectors whose spatial components should be investigated in the x, y coordinate system or relative to anatomical heart chamber coordinates, i.e. as the above discussed components: longitudinal, circular and radial, which very precisely reflect the characteristics of left ventricular myocardial mechanics. Similar principles apply to the analysis of strain and strain rate, which are also characterized by a change in the shape and location of selected myocardial areas throughout the cardiac cycle. The superiority of S and SR over D and V results from the elimination of the impact of the so called translation movements due to the dislocation of the whole organ inside the chest, and thus a change in the position of the heart with respect to the observation point – ultrasound probe, on these parameters. The inability to differentiate between active strain (caused by active contraction or relaxation) and passive strain (e.g. as a result of stretching of ventricular portion showing contractile inactivity by segments with maintained systolic function) remains a gap in the analysis of strain. Strain can be considered in relation to the individual segments (segmental strain), layers and, finally, the whole ventricle (global strain), which is calculated by averaging the values of strain in the individual segments(.

The term LV rotation refers to a rotational movement of the left ventricle around its long axis; the value of rotation is expressed in degrees. Normally, the apex and the basis of the left ventricle rotate in opposite directions. The total range of apical and basal motion is referred to as the twist angle and also expressed in degrees. The term torsion refers to the gradient in rotation along the long LV axis and is expressed in degrees per cm (Dg/cm) (Fig. 6).

Fig. 6

The torsion angle expresses the total range of the opposing LV apical and basal motion as well as the difference (expressed in degrees) between the positions of both these segments at specific time points. As can be seen from the figure, the graph illustrating the clockwise basal left ventricle rotation is violet-coded, the opposing apical rotation in blue-coded, and the torsion angle graph is white. In this case, the maximum torsion angle is 14° (an arrow)

Performing the examination

After opening the optimal view, the endocardial surface of the analyzed ventricle is outlined manually, and the outline of the epicardial surface is overlaid automatically, by the system, thus forming an area intended for analysis (ROI). After manual adjustment in the width and shape of the ROI, it is automatically divided by the system into six segments corresponding to LV segments. This is followed by generation of strain curves for each of the selected segments. These curves show the strain values for the individual segments as well as global strain values calculated by averaging all the individual values. If the analysis involves all three apical views, the system shows the movement of 17 segments on a bull’s-eye display( (Fig. 7).

Fig. 7

A simultaneous presentation of the peak longitudinal strain values on a Bull’s Eye diagram of all left ventricular segments is a convenient way of graphical representation of systolic function. The diagram is obtained as a result of the analysis of basic apical views: four-chamber, twochamber and left ventricular long axis view. The diagram is presented in the form of a color-coded map for all segments with the values of peak systolic strain of each segment. Additionally, global strain values for each view and global strain of the whole ventricle, are provided. It is also possible to generate similar diagrams for other investigated parameters

It is also possible to directly measure the time-to-peak strain and the postsystolic index (the percentage of postsystolic strain compared to the peak strain of each evaluated segment) (Figs. 8 and 9). Both these parameters are useful for further analysis – for an identification and quantitative assessment of ischemic regions or regions showing asynchrony due to other reasons.

Fig. 8

The primary objective of STE is an assessment of mechanical events in different functional regions of the heart. The method allows to identify differences at time when different segments achieve the maximum strain (i.e. systole), thus providing an insight into this mechanical aspect crucial for the global mechanical function, as well as provides information on the possible uneven blood supply in the individual segments or impaired spreading of the electrical excitation throughout the working muscle. In addition to typical strain graphs, a diagram with time points at which maximum strain (time to peak longitudinal strain) was achieved in the individual segments, is also presented. Minimal variation in this parameter (318–334 ms), indicating normal excitation and ventricular contraction, may be seen

Fig. 9

The postsystolic index, i.e. the percentage of postsystolic strain (following systole) compared to the peak strain of each evaluated segment, is of similar importance. In the presented figure, the postsystolic index is 0 for each segment, which indicates normal mechanical function of the investigated ventricle.

Strain curve analysis

Heart beat is a cyclical process, therefore it seems necessary to identify the beginning of the cycle, which will be the point of reference for the measurements of changes in the length of the evaluated segments. Optimally, it should be a moment of maximum stretch of the muscle fibers, i.e. end-diastole, however, it is difficult to select an image frame that adequately corresponds to this event in echocardiographic evaluation. Usually, a moment (image frame) preceding complete closure of the mitral valve is selected as a reference point. Alternatively, the beginning of the QRS complex, the peak of R-wave with the largest size of the left ventricle and, finally, when the peak (the highest positive value on the curve) of the longitudinal global strain occurs, can be used as a reference point. None of these alternatives is optimal due to significant time differences that may occur between each of these time points in different clinical situations, such as intraventricular blocks, segmental contractility impairment, etc. In the case of combined analysis of several views, especially in the case of variable heart rate, the duration of the individual cycles can vary substantially, thus additionally hindering an optimal choice of the baseline. Therefore, images recorded at the same heart rate should be chosen for analysis, if possible( (Fig. 10). Time point at which muscle fibers are maximally extended, i.e. the longest, serves as a reference point for strain measurements, therefore the systolic S observed during the examination due to fiber shortening has negative values.

Fig. 10

The figure shows a dot plot of the LV longitudinal global strain. The system accepted the R-wave as the beginning of the cycle. The point 3 on the curve indicates the point of end systolic strain (ESS), and is slightly preceded by point 2 – peak systolic strain (PSS), as well as point 1 – minor positive presystolic strain (PPS). This phenomenon illustrates the importance of the proper choice of cardiac cycle onset – if QRS complex was chosen as this point, there would be no positive deflections in the curve. Furthermore, there is no postsystolic strain (PSS) on the curve

The end-systole is another time point requiring precise determination as this is when maximum myocardial strain, i.e. the maximum S value, is expected. This point is determined by aortic valve closure (AVC), which may be identified using pulsed Doppler, by recording several images in the longitudinal parasternal or apical view; the point of maximum GS curve deflection can also be used.

The following important events can be identified during the analysis of strain curve:

end-systolic strain, (ESS) – the S value at a point considered to be the end of systole (aortic valve closure);

peak systolic strain (PSS) – the highest S value during systole; may occur at a time point other than ESS;

peak positive strain, (PPS) – minor local myocardial stretching can occur during the phase of early systole in a healthy heart or as a manifestation of regional dysfunction;

peak strain (PS) – the highest S value, regardless of the phase of the cycle in which it occurs; its occurrence should be documented;

post-systolic strain (PSS) – further ventricular deformation after aortic valve closure; minor PSS can indicate the above described early-diasystolic asynchrony, major PSS can indicate impaired segmental contractility or dyssynchrony of larger areas.

Determination of the time between the beginning of the cycle and the time point at which the above described events occur for each of the evaluated segments allows for a quantitative assessment of the differences in strain dynamics in certain myocardial regions(. Although the most efficient STE assessment is achieved for the thick-walled left ventricle, evaluation of other heart chambers, i.e. the right ventricle and both atria, is also possible under favorable conditions(.

Clinical applications

STE allows for a multifaceted insight into the systolic and diastolic myocardial function in a variety of physiological and pathological states, thus significantly extending the diagnostic possibilities of the methods used so far. For example, although the ejection fraction calculated based on the analysis of longitudinal strain components shows a good correlation with other techniques, the exceptional value of STE is the possibility of quantitative analysis of the individual segments allowing to detect early systolic dysfunction in the period when ejection fraction is still maintained(.

Hypertension

Gradual deterioration of longitudinal and radial strain components with maintained normal circular and twisting components, which enable compensation and the maintenance of normal global systolic function, is initially observed in hypertension and progressive concentric hypertrophy. STE allows to capture the sequence of these changes in the mechanical function of the ventricle in the period preceding the occurrence of significant systolic dysfunction(.

Ischemic heart disease

Decreased longitudinal strain values are observed in patients who have not yet developed segmental disorders; decreased S is a predictive factor of ischemic cardiomyopathy. Correlation was also shown between reduced global strain values and the level of indicator enzymes and the extent of necrosis in patients in the acute phase of myocardial infarction. The value of longitudinal strain measured early after reperfusion also proved to be a prognostic factor for post-infarction remodeling as well as adverse sequelae of myocardial infarction, such as congestive heart failure and death. Finally, it was shown that the longitudinal strain correlates with the extent of necrosis ([non]transmural myocardial infarction, the number of the segments involved, extent of post-infarction scarring), assessed using MRI. The values of the individual S components indicating an improvement in myocardial function due to revascularization were determined. Furthermore, the types of postsystolic mobility (after aortic valve closure), characterized by ischemic regional systolic dysfunction of the myocardium, were identified(.

Valvular heart defects

A limited increase in the S value during cardiac stress test in asymptomatic patients with mitral regurgitation proved to be a prognostic factor for postoperative myocardial dysfunction(. On the other hand, a very rapid circular and radial strain normalization observed after aortic valve surgeries, is an evidence of a very high dependence of these parameters on preload and afterload conditions(.

Heart failure

Gradual decrease in the global values of the longitudinal strain component along with increased severity of heart failure classified in accordance with the NYHA, was shown. Impairment of the circular and radial component is usually observed in the later period, in patients with class III – IV heart failure(. There is a significant impairment in the parameters characterizing left ventricular twist during heart failure. They increase in the phase of mild functional impairment, with apparent gradual normalization in the progression of diastolic dysfunction, to later decrease. Although no reduction in the range of the twist is observed in the initial phase of heart failure, a delay in the onset of untwisting occurs already in the earliest period, which is particularly manifested during physical exercise(. It is not fully understood if the increase in the twist parameters at the initial stages of diastolic failure is a mechanism compensating for the impaired relaxation or a consequence of reduced ventricular filling in early diastolic dysfunction. It was shown that the global circular component was a predictive factor for cardiovascular events in patients with heart failure and reduced ejection fraction(. Furthermore, it was observed, that the longitudinal strain component was a better prognostic factor for the course of disease compared to ejection fraction(.

Ventricular dyssynchrony

The phenomenon involving the loss of an appropriate sequence of ventricular contraction as well as individual ventricular components, referred to as dyssynchrony, often occurs in patients with heart failure and is considered to be an indicator of significant disease progression and poor prognosis. In the case of left bundle branch block or right ventricular stimulation, septal stimulation first occurs, resulting in a simultaneous stretching of free, unstimulated left ventricular wall, which reduces both the duration of diastole and the peak rate of pressure rise (dP/dtmax) during the isovolumetric contraction phase. A delayed contraction of the LV lateral wall dissipates the forces generated within the relaxing septum, thus decreasing the cardiac output. Uncoordinated contraction of the papillary muscles may additionally increase LV dysfunction due to developing mitral regurgitation. Dyssynchronous relaxation prolongs isovolumetric contraction, and thus additionally reduces LV filling.

There are multiple methods for echocardiographic imaging of left ventricular dyssynchrony (M-mode, STE, DTI and 3D)(. Establishing indications for resynchronization therapy and prediction of its efficacy is one of the main diagnostic goals. Presently, patient eligibility for cardiac resynchronization is based on clinical symptoms (cardiac insufficiency class III and IV according to NYHA), left ventricular function (EF <35%) and QRS duration in ECG (0.18 sec). However, despite the use of the above mentioned criteria, about onethird of patients undergoing cardiac resynchronization therapy do not respond to treatment, i.e. do not show improvement in the left ventricular function, which indicates the need to develop better qualification criteria. STE allows for an assessment of cardiac cycle subperiods and their differentiation in various ventricular regions and, particularly, it allows to demonstrate that there is a significant difference between the activation of the basal segment of the free right ventricular wall and the latest stimulated right ventricular segment(.

Cardiomyopathies

Characteristic impairment in all strain components is observed in patients with hypertrophic cardiomyopathy with preserved EF. Therefore the method is used for differentiation between cardiomyopathy and athlete’s heart as well as for monitoring of the course of the disease(. Characteristic abnormalities are also observed in patients with other forms of cardiomyopathy (ectatic cardiomyopathy)( and left ventricular non-compaction(.

Limitations of 2D STE

Speckles can be correctly observed provided that optimal myocardial visualization is achieved. Artifacts, such as acoustic shadows or reverberations can imitate or distort the image of speckles, and thus result in underestimation of the actual degree of strain. Therefore, if strain curves seem non-physiological, an insufficient quality of signal should be taken into account and the position of the ROI adjusted or the acquisition of suboptimal echocardiogram views should be repeated. Tracing algorithms used by the system smooth out the images and use a priori assumptions of normal ventricular function, which may lead to a misdiagnosis of regional dysfunction or affect the measurements in the adjacent segments(. The use of STE for the assessment of LV rotation may be limited by the quality of the imaging of the basal segment in the short axis. This is partly due to acoustic problems associated with the deep location of the basal portion of the ventricle and, partially, due to a large sector width necessary for full visualization of this structure. The measurements are additionally hampered by partial movement of the image beyond the plane of the beam due to the lowering of the AV valve rings towards the apex during contraction. Since the degree of left ventricular rotation increases towards the apex, it is important to optimize the apical view in the short axis. The assessment of global strain may be inadequate if too many segments were excluded from the analysis due to insufficient image quality. This is particularly important in the case of local contractility impairment, when the S values are unevenly distributed(.

STE evaluates movement with respect to a constant external observation point (probe). The advantage of this method is that it measures movement in any direction on the imaging plane, whereas DTI is limited to the measurement of a component facing the probe. This property allows for the measurements of circular and radial components, regardless of the direction of the beam. It should be noted, however, that STE is not completely free of angular dependence as the resolution of ultrasound images along the beam is better compared to transverse direction(. Therefore, STE shows the highest precision for the analysis of movement whose direction is in line with the direction of the ultrasonic beam. As in other 2D techniques, STE accuracy depends on the quality of the image. The analysis is based on an assumption that the examined structures can be traced continuously in subsequent image frames. This assumption is often difficult to be put to practice during the whole cardiac cycle due to a significant range of heart movement in the chest(.

A variety of systems for the procedure, and thus the inability to directly compare results obtained using devices from different manufacturers, represents a significant limitation. It is also one of the reasons for the difficulties in developing standards and values for different physiological states(. Undoubtedly, the largest advantage of this method, i.e. integration of multiple aspects of cardiac mechanics in a single test, is a major challenge. Regardless of the difficulties and limitations, the new method offers an expanded insight into the mechanical cardiac function in patients after cardiac surgeries, both in the immediate postoperative period, when e.g. impairment of the longitudinal component of the right ventricular systolic function is observed, as well as during distant postoperative follow up(. Further clinical research is needed to determine the actual value of the method in these applications. It is also desirable to standardize the technical aspects of the procedure performed using systems from different manufacturers as this will allow for a comparison of results obtained in different centers, which is currently impossible(.