SPECKLE TRACKING ECHOCARDIOGRAPHY : BASIC CONCEPTS AND CLINICAL APPLICATIONS

Hypertension being the most prevalent clinical entity in cardiology practice needs the most attention, as sub-clinical LV dysfunction ensues early on much before any decline in ejection fraction, as shown by Amal Ayoub et al16 in their study of 90 patients. Hypertensive patients with preserved ejection fraction when compared with a normal cohort showed sub-clinical dysfunction in a significantly higher number of patients, as shown in Table 5. In this regard it has been found that impairment of longitudinal and radial strain precedes the reduction of circumferential strain and torsion, thus, preserving the LVEF.


INTRODUCTION:
The heart, as we know, is a mechanical organ, the primary function of which is to pump an effective amount of blood to meet the requirements of our body. This mechanical function is most commonly measured by echocardiography and expressed in numbers of various units. The most important parameter in this regard is Ejection Fraction which is calculated for every patient. This is the difference in diastolic and systolic volumes of left ventricle per diastolic volume and expressed as a percentage. The parameter which gets the most attention in an Echo report is EF, but is this really that robust? If the same expression is used taking diameters as dimensions, we get Fractional Shortening, a parameter less recognized than EF but, still applicable. Other parameters of mechanical performance like Circumferential Fiber Shortening (Vcf). Mitral Annular Plane Systolic Excursion (MAPSE) and E-point Septal Separation (EPSS) receive even less attention, figure 1.
Estimation of EF has the following drawbacks: • Eyeballing is frequently employed in its estimation.
• Quantification methods are cumbersome.
• Tracing of endocardial borders for volume assessment is not possible in many patients and interpolation can be erroneous.
• Variability in measurements, both inter and intra-observer, is too much.
• A compromised EF usually signifies a point of no return.
• Limited or no assessment of radial and circumferential strains as well as of twist and torsion.
So, another parameter is needed for assessment of cardiac (mainly LV) function; 'strain' is such a parameter. It has been defined as "the ratio of change in length (∆L) to resting length (Lo) upon application of force to a muscle", expressed as percentage and symbolized by Epsilon, figure  2.
Victoria Delgado et al in their article 'imaging in 2018' published in European Heart Journal ascertained the importance of strain over 2D EF in the following words, "In 2018, strain imaging with echocardiography has provided important patho-physiological insights in various cardiovascular diseases and the evidence demonstrating its incremental prognostic value over left ventricular ejection fraction is growing."1 Kalam K, in his article on the prognostic implications of

FIG. 2 DEFINITION OF STRAIN
global LV dysfunction, mentions that, "as compared to LVEF, GLS has superior prognostic value for predicting all-cause mortality, cardiac death, malignant arrhythmia, hospitalization due to heart failure, urgent valve surgery, or heart transplantation and acute coronary ischemic event". 2

LITERATURE SEARCH:
Google and PubMed search with the words 'speckle tracking' and 'basic concepts' and 'clinical applications' yielded 384 citations. By manual search, 37 articles were found pertinent, from which material for this manuscript has been extracted.

TEXT:
To understand it better, we have to look into the basics of cardiac muscle architecture and the mechanics of cardiac contraction.

CARDIAC MYOCYTE ARCHITECTURE:
Left ventricular muscle fibers, as shown in figure 3 are arranged in a helical fashion with the sub-endocardial fibers forming a right handed helix. As these come in mid layers, they get a more horizontal orientation and further on as they become sub-epicardial form a left handed helix 3 . The sub-endocardial fibers are more vertically aligned so they are responsible for the longitudinal shortening of left ventricle whereas the fibers of the other two layers cause contraction in radial and circumferential directions.

BASIC CARDIAC MECHANICS:
Left ventricular contraction, as mentioned above, occurs in three different directions, figure4, as per orientation of muscle fibers4. The longitudinal fibers of sub-endocardial region produce longitudinal shortening, hence, are responsible for deformation in this direction called "longitudinal strain" whereas deformation in radial and circumferential directions is known as "circumferential" and "radial" strain respectively.
Myocardial strain can thus be evaluated along three axes according to cardiac muscle deformation. Electromechanical activation would result in a shortening in the longitudinal and circumferential directions (resulting in a negative strain) and a thickening in the radial direction (resulting in a positive Pak Heart J 2020 Vol. 53 (01) : 10 -23 The apex and base of the heart show a rotational movement in opposite directions with the base in a clock-wise direction (about 10 degrees normally) and the apex in a counter-clockwise direction (about 5 degrees normally). The difference of these two rotations is known as "TWIST", and is responsible for squeezing blood out of the left ventricle. Due to all these motions the base of LV moves towards the apex during systole and thus shortens in length, TWIST divided by this shortening of LV length is known as "TOR-SION".Twist and Torsion movements produce a wriggling effect. All these movements ensure that a 15% decrease in the length of myocyte results in an Ejection Fraction of 60%.

HOW TO MEASURE STRAIN:
To measure deformation or strain various modalities have been used in research and clinical practice. The gold standard in this regard is "Sono-micrometery" in which crystals are attached to the surface of the left ventricle which shorten their distance during systole and this parameter is measured sonically. This can only be applied in research and obviously cannot be used in routine practice.
Another gold standard method is MRI Tagging. Due to its high cost and non-availability as a routine procedure, this is also not very useful.
Doppler, as we know, is a differential effect due to the direction of movement of an object towards or away from the recording source (as initially shown by Christian Andreas Doppler, during his study of the stars). This phenomenon has very aptly been applied in Echocardiography for decades now to record velocity of blood flow in different directions for quantification of stenotic, regurgitant and shunt blood flow. However, blood, as compared to tissues, moves at a very fast pace. To measure the velocity of tissue effective filters are incorporated in Echo machines to filter off higher velocities and to show only the tissue velocities which at times are color-coded as well. This is known as "Tissue Doppler Imaging" or "TDI", figure 5.
For deformation imaging, this has been used extensively but because of the following reasons has gone out of favor: • TDI is angle dependent.
• Tissue movement is assessed in relation to transducer position.
• Movement can be assessed in one direction only.
• Spatial resolution is limited.
The latest addition to our armamentarium for the assessment of strain is: "SPECKLE TRACKING".

BASICS OF SPECKLE TRACKING:
Ultrasonic images are created by reflection of sound waves by targets at least half the size of the wavelength, rest of it is refracted or back-scattered. 'Speckles' are produced by the constructive and destructive interference of ultrasound backscattered from structures smaller than the wave length.
If an M-mode cursor is applied, 'lines' would be created as 'Finger Prints', figure 6.
In speckle tracking we track the backscatter created by ultrasound comprising of 20 to 25 pixels, frame by frame, thus, comparing the velocities of two points in a given segment which gives the strain rate from which strain value can be derived easily. It is non-Doppler and independent of the angle of insonation. It holds promise to reduce inter-observer and intra-observer variability in assessing regional LV function and to improve patient care by identifying sub-clinical disease. 6 13 Pak Heart J 2020 Vol. 53 (01) : 10 -23 Speckle Tracking Echocardiography: Basic Concepts and Clinical Applications

FIG. 5: TISSUE DOPPLER IMAGING
of tissue effective filters are incorporated in Echo machines to filter off higher velocities and to show only the tissue velocities which at times are color-coded as well. This is known as "Tissue Doppler Imaging" or "TDI", figure 5.

ACQUISTION AND ANALYSIS OF DATA:
To standardize strain imaging, European Association of Echocardiography and the American Society of Echocardiography along with technical representatives from all interested vendors made a concerted effort to reduce inter-vendor variability of strain measurement and the three of them prepared a technical document which provided definitions, names, abbreviations, formulae, and procedures for calculation of physical quantities derived from speckle tracking echocardiography to create a common standard. 7 The foremost consideration for data acquisition is the frame rate, which should ideally be between 60-100 fps.

. ACQUISITION OF DATA WITH MARKING OF REGION OF INTEREST, THE BAR BELOW WILL SHOW WHETHER THE ROI HAS BEEN MARKED CORRECTLY WITH BUTTON FOR APPROVAL
image is good and both endocardial and epicardial borders are clearly seen.
From apical position acquire the three images i.e. 3, 4 and 2 chamber views, as shown in figure7, in that order and store the clips because assessment can only be done on stored images, the requirements of frame rate, image resolution and heart rate should be the same for all views. Define the end systolic frame, by Aortic valve closure, as a landmark easily visible in this view. Other methods for it are either the end of T wave or from Doppler of aortic valve with its closing sign as a landmark.
ROI of each view is drawn and adjustments are made if needed, the bar below will show if the system considers movement of speckles satisfactory, figure 8, there is a provision that if the system cannot recognize the movements as satisfactory for a given segment but the observer thinks it satisfactory he can change the X to √ sign and when OK is clicked the software reads the segmental values and ROI overlying the 2D image would be seen in shades of blue and red.
Four images are obtained as follows: It has been shown in figure 9 (upper left ) that the region of interest is laid over a 2D image with the segments shown in particular colors; ROI should be so adjusted to encompass the whole myocardium, "Parametric view". The picture in the lower left corner shows the strain values of individual segments overlaid over a 2D image, "Peak systolic strain.
The upper right corner image shows the strain changes of different segments in one view (here 2C) in a graphical pattern of different colors corresponding to the color of segments of parametric view. The peak value is marked by a circle and the average of all the six segments is represented by a white dotted line. The picture in the lower right corner "ribbon graph", shows the change of strain (according to color) during one complete cardiac cycle. The ribbons are marked by the same colors with which individual segments have been colored in the first picture, so that one ribbon shows the complete strain changes in one cardiac cycle of a particular segment. Arrangement of ribbons is: basal inferior, mid-inferior, apical inferior, apical anterior, mid-anterior and basal anterior (in that order from top to bottom of the picture). With ECG, phase of cardiac cycle can easily be correlated.

FIG. 10 BULL'S EYE VIEW
Demarcation of LV walls is as per ASE guidelines for LV regional function assessment.
Since no clear-cut guidelines are available for demarcation of normal limits of LV strain, three references can be cited in this regard until the leading echocardiographic societies come to a conclusion, as follows in Table 1.

RIGHT VENTRICULAR SPECKLE TRACKING:
RV strain assessment is important in diseases leading to RV dysfunction, such as congenital anomalies (e.g. Tetralogy of Fallot, arrhythmogenic RV dysplasia, pulmonary arterial hypertension, and pulmonary thromboembolism).
For analysis of right ventricle by speckle tracking, an RV-focused apical four-chamber view is obtained and the region of interest (ROI) is drawn encompassing RV free wall and inferior septum. Each wall is divided into three equal segments, figure 11. The rest of the procedure is the same as for LV assessment. However, the cut-off values for normal RV strain are much higher than those for LV.   10 19.8% -25.2%

LEFT ATRIAL SPECKLE TRACKING:
Left atrium functions as a reservoir, conduit, and pump. It is tracked in apical four-chamber view so that the region of interest (ROI) covers the inter-atrial septum, superior wall and lateral wall.
Atrial strain is evaluated in various conditions, such as hypertension, diabetes, heart failure, ischemic and valvular heart disease, atrial fibrillation, including the facilitation of stroke risk calculation and for prognostic purposes. Faraz Pathan et al. 14 have shown the normal strain values as shown in Table 4.

RIGHT ATRIAL SPECKLE TRACKING:
Right atrium is said to be a neglected chamber. An assessment of it by speckle tracking is particularly difficult because of its thin walls, contour, many structures fitted into a small space and it can be imaged only in one view i.e. four chamber apical view. However, atrial mechanics involve the same three phases as in left atrium, i.e. reservoir, conduit and pump. Aitzaz Bin Sultan Rai et al. 15 , have emphasized the importance of right atrial strain assessment in the following conditions: •

CLINICAL INDICATIONS FOR SPECKLE TRACKING ECHOCARDIOGRAPHY:
Speckle tracking echocardiography is gaining immense value in many clinical situations because of its robustness, ease of use and off-line assessment. Clinical indications are:

HYPERTENSION:
Hypertension being the most prevalent clinical entity in cardiology practice needs the most attention, as sub-clinical LV dysfunction ensues early on much before any decline in ejection fraction, as shown by Amal Ayoub et al 16 in their study of 90 patients. Hypertensive patients with preserved ejection fraction when compared with a normal cohort showed sub-clinical dysfunction in a significantly higher number of patients, as shown in Table 5. In this regard it has been found that impairment of longitudinal and radial strain precedes the reduction of circumferential strain and torsion, thus, preserving the LVEF.

HEART FAILURE:
Heart failure is a very common cause of hospital admissions. To see the reasons for these readmissions clinical, echo and strain parameters were compared.
Causes of re-admission are heterogeneous. Clinical parameters for assessment of risk for re-admission are modest, echo parameters are promising but the most commonly used parameter, "EF", is inconsistent in this regard. GLS is associated with HF readmission, independent Pak Heart J 2020 Vol. 53 (01) : 10 -23

VALVULAR HEART DISEASES:
Strain analysis has been found to be of value in this group of patients as it detects sub-clinical LV dysfunction in patients who have not yet developed symptoms or have dropped their ejection fractions. An earlier intervention will ensure better post intervention results. Aortic Stenosis leads to maladaptive LV remodeling that includes sub-endocardial ischemia and fibrosis. Sub-endocardial myofiber function dictates LV longitudinal contraction. Therefore, with advanced AS, it is not surprising that longitudinal strain is impaired.
Joanna Luszczak et al. 20  GLS differs significantly depending on the severity of valve stenosis, with values of 17.1%, 16.4%, and 14.5% LV systolic shortening for mild, moderate, and severe stenosis respectively, despite a maintained LVEF. In addition to GLS, endocardial radial strain is also found to be reduced in proportion to severity of disease; however, epicardial radial strain and circumferential strain is preserved in asymptomatic AS. 21 In Low-flow Low-gradient severe AS, longitudinal strain is found to decrease even further compared with that in individuals with high-gradient AS (11.6% vs. 14.8% LV shortening), suggesting that progressive longitudinal dysfunction may be a contributing factor to the low-flow AS phenomenon. These observations suggest that in patients with asymptomatic severe AS, decreases in longitudinal strain can identify high-risk candidates for aortic valve replacement (AVR).
At present, there is no clear consensus regarding minimum changes or absolute GLS that should warrant earlier intervention for AS treatment. However, it has been shown that in patients with normal or depressed LVEF undergoing trans-catheter AVR or surgical AVR, those with impaired GLS had a significantly higher risk of cardiac morbidity and death.
Following definitive procedural intervention, significant reductions in trans-aortic gradients are observed with associated improvements in LV strain patterns. Following either surgical or trans-catheter AVR, significant increases in longitudinal strain have been observed.
The prognostic value of LV GLS was also demonstrated by Arnold C. et al. in 294 patients with severe aortic stenosis. Patients were divided according to a cut-off value of LV GLS of -14%. Patients with more preserved LV GLS (<-14%) had better survival when compared with patients with more impaired LV GLS (>-14%). Among patients with more impaired LV GLS, there was no significant difference in survival when patients were subdivided according to LVEF (>55% vs. <55%). Each 1% impairment in LV GLS was independently associated with 17 % increased risk of all-cause mortality. 22 The range of cutoffs observed for GLS that appear to show improved outcome after AVR in normal LVEF patients are broad, and range from 12.1% to 17.8% LV shortening. These observations suggest that in patients with severe AS and a declining GLS, early AVR should be considered. 23 Similar to Aortic stenosis, STE helps in Mitral regurgitation to select patients for surgery and predict post-operative results. As after-load is reduced in this condition, ejection fraction as a marker of LV function is very deceptive. Lancelloti P et al 24 have shown in their study that in patients with severe MR and normal LVEF, those who required surgery during the follow-up had a lower GLS.
Magne et al 25 in their study have shown that in patients with mitral regurgitation who fail to improve their GLS by 2%, on exercise, will have adverse cardiac events.

ISCHEMIC HEART DISEASE:
To evaluate ischemia, many parameters have been elaborated and compared with traditional methods like stress echocardiography, perfusion imaging and coronary angiography with good correlation in various studies. The parameters of STE found to indicate ischemia are: • Global peak longitudinal strain

• Dispersion of longitudinal strain in various segments
• Post-systolic stress (PSS) and PSS index • Early systolic lengthening Wei Chaun et al 26 , in their study of 152 pts for the assessment of CAD by GLS found that GLS was decreased in CAD, LSD (difference between segment with highest and that with lowest strain) was increased, whereas, the ratio of these two was also higher in CAD patients. Characterization of severe CAD (LM/3VD) and less severe disease (2VD/1VD) by STE has also been reported.
Ischemic segments show the features of post systolic shortening in which the myocardial segment continues to shorten even when systole has ended and this goes on into diastole as PSS figure 12. Philip Brainan group has reported its significance in their article. 27 Early systolic lengthening, figure 13, is a feature of ischemic myocardium in which the contraction in ischemic segments lag behind the electrical events so that the pressure build-up by contraction of the normal segment increases the length of ischemic segments resulting in early systolic lengthening noted on STE. The duration of this corresponds with the extent of ischemia, as exemplified by Marit et al in their study. 28 Kristina and colleagues have demonstrated that patients who have sustained an MI can be predicted to have arrhythmia on the basis of mechanical dispersion, i.e. the difference in time required for the segmental graphs to reach peak strain 29 , a value of 75 msec is indicative of high propensity of ventricular arrhythmias in this patient group, making them candidates for ICD. Evaluation of chest pain being ischemic or otherwise keeps most of the hospital ER busy and packed with patients, as this requires a lot of investigations as shown in Table 6 The TARGET 33 and STARTER 34 studies have shown that radial strain assessment helps in locating scar tissue and selection of optimal place for lead placement.
The parameters used for dys-synchrony assessment are: A. Maximum time delay between peak systolic strain of two segments (usually between the antero-septal and postero-lateral walls, DT) B.
Dys-synchrony index of LV 35 , taken from the standard deviation of time to peak systolic strain.

CARDIO-ONCOLOGY:
This is a rapidly growing field for the application of STE as most chemotherapeutic agents cause myocardial dysfunction by various mechanisms. Knowledge of sub-clinical dysfunction in these patients helps in altering the regimen which is more toxic and institution of cardiac drugs which can ameliorate the detrimental effect of these 20 Pak Heart J 2020 Vol. 53 (01) : 10 -23

FIG. 14 TYPICAL AND ATYPICAL LBBB SYNCHRONOUS (NON-RESPONDERS) AND ASYNCHRONOUS (RESPONDERS)
agents. There is a growing institution of clinics in hospitals which deal with oncology patients where cardiac and oncology consultations go side by side.
A King et al 36 have demonstrated the utility of GLS in their study of 43 patients of breast cancer (HER 2 positive) on Herceptin treatment. They showed that GLS assessment is better than 2D EF assessment in such patients as it has more intra and inter-observer variation, which results in a reduced assessment of EF rather than a real clinical change. They concluded that analysis by direct comparison, intra class correlation (ICC) and co-efficient of variation (CV) and Bland-Altman plots demonstrated that GLS is a more reproducible measurement than 2D EF.
Isaac B Rhea et al 37 in their study of 120 cancer patients (of various types) with normal ejection fraction analyzed 17 clinical and 6 echocardiographic parameters along with GLS for prognostic assessment with regard to mortality. They found that only ECOGP (Eastern Co-operative group oncology performance), male sex and GLS were significantly associated with mortality. Adding GLS to significant clinical variables provided incremental prognostic information.

CAVEATS:
Being an evolving modality, it has many caveats which need to be covered up with further research. Some are: • Sinus rhythm is mandatory, non-sinus rhythms like atrial fibrillation, frequent ventricular ectopics and paced patients cannot be analyzed.
• Imaging needs to be of good quality.
• Vendor differences have not been sorted out completely.
• No guidelines from any agency have yet defined the clear limits of normal and abnormal.

CONCLUSION:
With the sophistication of Speckle tracking echocardiography sub-clinical cardiac dysfunction (both global and regional) can be detected by utilizing 2-D images (not needing Doppler) much before conventional parameters show any deterioration. Furthermore STE helps in the assessment of rotational and torsional dynamics, and thus has huge potential in numerous fields of cardiology.

ACKNOWLEDGMENTS:
Thanks are due to my daughter Ms. Huffsa Imran in typing and proof reading of this manuscript.