• 2-dimensional strain;
  • Cardiology;
  • Contractility;
  • Echocardiography;
  • Myocardial function


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgment
  8. References
  9. Supporting Information

Background: The quantification of equine left ventricular (LV) function is generally limited to short-axis M-mode measurements. However, LV deformation is 3-dimensional (3D) and consists of longitudinal shortening, circumferential shortening, and radial thickening. In human medicine, longitudinal motion is the best marker of subtle myocardial dysfunction.

Objectives: To evaluate the feasibility and reliability of 2-dimensional speckle tracking (2DST) for quantifying equine LV longitudinal function.

Animals: Ten healthy untrained trotter horses; 9.6 ± 4.4 years; 509 ± 58 kg.

Methods: Prospective study. Repeated echocardiographic examinations were performed by 2 observers from a modified 4-chamber view. Global, segmental, and averaged peak values and timing of longitudinal strain (SL), strain rate (SrL), velocity (VL), and displacement (DL) were measured in 4 LV wall segments. The inter- and intraobserver within- and between-day variability was assessed by calculating the coefficients of variation for repeated measurements.

Results: 2DST analysis was feasible in each exam. The variability of peak systolic values and peak timing was low to moderate, whereas peak diastolic values showed a higher variability. Significant segmental differences were demonstrated. DL and VL presented a prominent base-to-midwall gradient. SL and SrL values were similar in all segments except the basal septal segment, which showed a significantly lower peak SL occurring about 60 ms later compared with the other segments.

Conclusions and Clinical Importance: 2DST is a reliable technique for measuring systolic LV longitudinal motion in healthy horses. This study provides preliminary reference values, which can be used when evaluating the technique in a clinical setting.


2-dimensional speckle tracking


time of aortic valve closure automatically calculated by the software


time of aortic valve closure measured by M-mode


peak systolic longitudinal displacement


frames per second


heart rate


left ventricle


region of interest


peak maximal longitudinal strain


peak positive longitudinal strain


peak systolic longitudinal strain


peak late diastolic longitudinal strain rate


peak early diastolic longitudinal strain rate


peak systolic longitudinal strain rate


tissue Doppler imaging


peak late diastolic longitudinal velocity


peak early diastolic longitudinal velocity


peak systolic longitudinal velocity

The quantification of myocardial function is important in horses, particularly in cases of poor performance. The objective evaluation of left ventricular (LV) function is traditionally based on M-mode measurements such as fractional shortening and wall thickening.1,2 These measurements provide a rather rough estimate of global LV function, as they are focal and 1-dimensional (1D). Recently, new techniques have been introduced in equine cardiology that require further investigation. Tissue Doppler imaging (TDI) has been used to measure the systolic and diastolic velocities of the myocardial walls.3–5 However, these measurements are easily influenced by the insonation angle and total heart motion. This limitation has led to a growing interest in strain imaging in human medicine. Strain is a measure of the amount of deformation of the myocardial walls, expressed in percent (%). Lengthening or thickening is indicated by a positive strain value, while shortening or thinning is negative. The rate of deformation is the strain rate, expressed in per seconds (s−1). TDI-based strain imaging is slightly less angle-dependent compared with velocity measurements.

Two-dimensional speckle tracking (2DST) is not angle-dependent and allows quantification of myocardial strain in the two dimensions of the ultrasound image based on tracking speckles in the myocardial walls.6,7 2DST has been reliably applied in short-axis images for quantification of equine radial and circumferential LV motion.8 However, LV contraction is a 3D movement consisting of radial thickening, circumferential shortening, and longitudinal shortening. In human medicine, longitudinal function is described as the best prognostic factor in several conditions such as valvular disease.9–11 Although 2DST has recently been applied to measure equine LV longitudinal function during stress-echocardiography, no data are available on the reliability of these measurements.12

The purpose of our study was to assess global and regional LV longitudinal function in healthy horses at rest. We hypothesized that 2DST methods allow quantification of LV longitudinal strain (SL), strain rate (SrL), velocity (VL), and displacement (DL) based on 2D grayscale images. The 2DST technique and the feasibility and reliability of the evaluation of equine LV longitudinal function are described.

Material and Methods

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgment
  8. References
  9. Supporting Information

Study Population

The study population consisted of 10 healthy untrained trotter horses (7 mares, 3 geldings) aged 9.6 ± 4.4 years (mean ± SD) with a body weight of 509 ± 58 kg. Animal handling and care was performed following the guidelines of the local ethical committee. A comprehensive examination was performed before the study to exclude cardiovascular or respiratory disease. This consisted of a general physical examination, thorough cardiac auscultation, 30 min ECG at rest and routine 2D, M-mode, and color Doppler echocardiography.


All horses were examined without sedation at heart rates (HR) below 45 beats per minute. Echocardiography was performed with an ultrasound unita with phased array transducerb at a frequency of 1.6/3.2 MHz (octave harmonics). A base-apex ECG was recorded simultaneously. At least 3 nonconsecutive cardiac cycles from each view were stored in cineloop format.

M-mode recordings of the aortic valve were made in the right parasternal LV outflow tract long-axis view. For the evaluation of LV longitudinal function, a standard right parasternal 4-chamber view1 was obtained. The imaging depth ranged from 26 to 28 cm depending on LV size, with a single focus positioned at 20 cm. The sector width was reduced to 55° in order to achieve a frame rate of at least 40 frames per second (fps). The view was then slightly modified to include as much of the LV as possible while the mitral annulus remained visible throughout the entire cardiac cycle. As a result, the apex was not always visualized (Fig 1).


Figure 1.  Modified 4-chamber view for 2DST analysis of LV longitudinal function. The ROI is automatically divided into 6 segments: “basSept,”“midSept,” and “apSept” for the interventricular septum; “basLat,”“midLat,” and “apLat” for the LV free wall. Tracking quality is verified by the software and displayed at the bottom of the screen. Approved segments are marked with a green “V.” The apical segments are rejected, as indicated by a red “X.”

Download figure to PowerPoint

Offline Analysis

Measurements were performed by a commercially available software package.c For each exam, 3 nonconsecutive cardiac cycles per view were analyzed one by one and the 3 measurements were averaged for all further analyses. The time of aortic valve closure (AVCMM) was measured manually as the time interval between the R-wave on the ECG and AVC in the M-mode image. The RR interval was recorded and instantaneous HR was calculated as 60,000/RR interval.

LV longitudinal function was assessed from the modified 4-chamber view with the “2D Strain” application, using the “4CH” mode. A U-shaped region of interest (ROI) was placed on the myocardium in a frame at end systole by tracing the endocardial border from the septal to the lateral wall insertion of the mitral valve, without following excessive bulging of the walls. This line was partially drawn outside the image where the apical endocardial border was expected. Afterwards, the ROI width was adjusted in order to cover the entire myocardium without including the epicardial borders. The software algorithm subdivided this ROI into 6 segments (Fig 1). The apical segments were excluded from further analysis as these segments were not visualized. The midwall segments were manually corrected if they did not fit the image width. Next, the program automatically selected acoustic speckles that were tracked throughout the cardiac cycle in longitudinal and transverse (radial) direction. Based on the tracking quality, the software approved or rejected the segments for further analysis. In addition, tracking quality was evaluated by visual assessment during motion playback. If necessary, the ROI was adjusted and tracking was repeated until adequate tracking quality was achieved. Segments were excluded if they were rejected by the software or if tracking was visually insufficient. If tracking was visually accurate during systole but not during diastole, only systolic measurements were included for further analysis. Occasionally, the last systolic frames were poorly tracked in some segments. As the effect of 1 or 2 poorly tracked frames was thought to be minimal, these segments were included for further analysis. However, this might have resulted in a small underestimation of strain values in these segments.

After tracking quality had been verified for all segments, the trace analysis screen was shown. The following curves were displayed: VL, SL, SrL, DL, and transverse displacement. The default settings for spatial smoothing, temporal smoothing, and drift compensation were used. The time of aortic valve closure (AVCa) was calculated by the software based on the SL curves of all LV segments. Segmental systolic (s), early diastolic (e), and late diastolic (a) peak values were indicated on the curves and could be manually adjusted if necessary. Global and averaged peak values were determined as measurements of global LV function. Global values were computed by the software from the whole ROI as a single segment. These values were not visible in the trace analysis screen but were included when all data were exported from the software. Averaged values were calculated manually for all measurements as the average of the segmental values. Furthermore, a distinction was made between peak systolic longitudinal (SLS) and maximal longitudinal (SLG) strain. SLS and SLG were the same if only 1 peak was present in the SL curve, occurring before AVCa. If the curve showed 2 peaks with the 2nd and highest peak after AVCa, this peak was considered SLG and the peak before AVCa was SLS. If only 1 peak was present, occurring after AVCa, the strain value at the time of AVC was considered SLS whereas the peak after AVCa was SLG.

Reliability of Echocardiographic Variables

The reliability of all 2DST measurements was evaluated by comparing repeated echocardiographic examinations and offline measurements performed by 2 experienced echocardiographers (A.D. and G.V.L.). In total, 30 exams were performed. All horses were examined by observer 1 (A.D.) on 2 separate days with an interval of 1 day. On one of these occasions (day 1 in 5 horses and day 2 in 5 horses), the echocardiographic examination was repeated by observer 2 (G.V.L.) immediately before (5 horses) or after (5 horses) the 1st observer. The offline analysis was first performed by observer 1 for all echocardiographic exams (n = 30). Next, 1 exam of each horse was measured by both observers on a 2nd, separate occasion to determine measurement variability (n = 10). At any time during offline analysis, both observers were blinded to echocardiographer, horse, day, and any previous results.

Statistical Analysis

Statistical analyses were performed by dedicated computer software.d Within- and between-day intra- and interobserver variability were determined based on the analysis of variance results of all measurements. The within-day interobserver variability was obtained by comparing the results of the exams performed on the same day by observer 1 and observer 2. The numerical values for the reported coefficients of variation (CV) were derived from the expressions of the expected mean square errors from a 2-factor (observer and horse) analysis of variance of the natural logarithms of the measurements. In analogy, the between-day intraobserver variability was determined by comparing the exams recorded on 2 separate days by observer 1 by a 2-factor (day and horse) analysis of variance. Similarly, the measurement variability was obtained by repeated measurements of 30 identical cycles (1 exam per horse) on 2 different days (intraobserver measurement variability) or by 2 different observers (interobserver measurement variability). The degree of variability was defined based on the CV as follows: CV < 15%, low variability; CV 15–25%, moderate variability; and CV > 25%, high variability.13

Summary statistics for the different measurements (mean ± SD; n = 10 horses) were calculated using the average measurement of 3 nonconsecutive cycles per horse. Peak values and timing in the different segments were compared by analysis of variance; posthoc tests were performed by the Bonferroni correction method. These data were displayed as box plots, with the box span indicating the middle half of the observations, the line in the box marking the median, and the whiskers indicating the range of observations. Extreme values were plotted individually at the end of the whiskers. Agreement between AVCa and AVCMM was assessed by Bland-Altman 95% limits of agreement. A paired t-test was used to compare both measurements of AVC and their corresponding HR, as well as to compare global and averaged values of SL and SrL. The level of significance was α= 0.05.


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgment
  8. References
  9. Supporting Information

2DST analysis was feasible on 3 cycles of each exam using the modified 4-chamber view at a frame rate of 41.1 fps. Figure 2 shows an example of the curves displayed by the software after approval of tracking quality. The shape of these curves was similar in all horses and peak values could be easily distinguished. All apical segments were excluded as they were not visualized. Tracking quality was approved in 574 of the remaining 600 segments (95.7%). However, the tracking process was visually less accurate in diastole. In 6.1% of the approved segments, tracking was very poor and diastolic values were excluded from further analysis. Tracking quality of transverse motion was often visually inadequate and the variability of the measurements was very high (CV > 25%). Therefore, these measurements are not further discussed.


Figure 2.  Examples of the curves in the “Results” screen after approval of tracking quality. On the left, a 2D grayscale image is displayed showing the tracked ROI. A parametric color-coded M-mode is depicted below. On the right, the segmental traces are displayed. The vertical axis represents values of the selected variable; the horizontal axis shows time (ms) and the ECG. AVCa is marked by a vertical green line. Peak values are automatically indicated on the curves and tabulated below. After verifying the correct position of these peaks, measurements can be approved in the column next to the table. (A) VL, longitudinal velocity; (B) DL, longitudinal displacement; (C) SL, longitudinal strain; (D) SrL, longitudinal strain rate; (E) DT, transverse displacement.

Download figure to PowerPoint

Tables 1 and 2 show the peak values of global, averaged, and segmental 2DST measurements in the study population. The global SLG value was not significantly different from the averaged SLG (P= .055). The global SrL values were slightly lower than the averaged SrL values (P < .001). Significant segmental differences were found. As illustrated in Figure 3A–B, DL and VL presented a remarkable base-to-midwall gradient (P < .001). Conversely, SL and SrL values were similar in all segments except the basal septal segment (Fig 3C–D). This segment showed a significantly less negative SLS (P < .001) and SLG (P= .022).

Table 1.   Globala values of 2DST measurements of LV longitudinal function in 10 healthy adult trotter horses.
VariableMeanSDIntraobserver Measurement Variability CV (%)Interobserver Measurement Variability CV (%)Between-Day Intraobserver Variability CV (%)Within-Day Interobserver Variability CV (%)
  • SD, standard deviation; CV, coefficient of variation; HR, heart rate; AVCa, time of aortic valve closure automatically calculated by the software; SLG, peak maximal longitudinal strain; SrLA, peak late diastolic longitudinal strain rate; SrLE, peak early diastolic longitudinal strain rate; SrLS, peak systolic longitudinal strain rate.

  • a

    Calculated by the software from the whole ROI as a single segment.

HR (bpm)33.43.883.
AVCa (ms)528263.
SLG (%)−24.31.727.27.48.310.1
SrLS (s−1)−0.880.
SrLE (s−1)0.980.0910.79.310.811.3
SrLA (s−1)0.630.1214.114.516.620.5
Table 2.   Segmental and averaged peak values of 2DST measurements of LV longitudinal function in 10 healthy adult trotter horses.
VariableSegmentanMeanSDIntraobserver Measurement Variability CV (%)Interobserver Measurement Variability CV (%)Between-Day Intraobserver Variability CV (%)Within-Day Interobserver Variability CV (%)
  • n, number of segments in which the variable could be determined (on a total of 150); SD, standard deviation; CV, coefficient of variation; NA, not available; Avg, average of the approved segments; DLS, peak systolic longitudinal displacement; SLG, peak maximal longitudinal strain; SLP, peak positive longitudinal strain; SLS, peak systolic longitudinal strain; SrLA, peak late diastolic longitudinal strain rate; SrLE, peak early diastolic longitudinal strain rate; SrLS, peak systolic longitudinal strain rate; VLA, peak late diastolic longitudinal velocity; VLE, peak early diastolic longitudinal velocity; VLS, peak systolic longitudinal velocity.

  • a

    Segments as illustrated in Figure 1.

SLG (%)Avg −24.802.367.
SLS (%)Avg −24.492.527.
SLP (%)Avg 1.800.4967.957.560.961.5
VLS (m/s)Avg 8.370.447.67.710.110.1
VLE (m/s)Avg −7.710.5010.310.511.911.9
VLA (m/s)Avg −7.460.769.09.211.511.7
SrLS (s−1)Avg −0.950.
SrLE (s−1)Avg
SrLA (s−1)Avg 0.670.1313.514.015.016.1
DLS (mm)Avg 37.605.1812.311.213.215.5

Figure 3.  Graphical illustration of segmental differences of 2DST measurements of LV longitudinal function. For each segment, the median and spread of the peak value or timing is indicated by a boxplot, based on the averaged measurement values for each horse (n= 10). Boxes marked with different letters are significantly different from each other. (A) VLS, peak systolic longitudinal velocity (cm/s); (B) DLS, peak systolic longitudinal displacement (mm); (C) SrLS, peak systolic longitudinal strain rate (s−1); (D) SLG, peak maximal longitudinal strain (%); (E) time to SLG, time to peak maximal longitudinal strain (ms); (F) time to SrLA, time to peak late diastolic longitudinal strain rate (ms).

Download figure to PowerPoint

The mean AVCa was 528 ms, which did not differ significantly from the AVCMM of 537 ms (P= .070). A mean bias of −10 ms with a standard deviation of 14 ms was present, resulting in 95% limits of agreement between −36 and +18 ms. The mean RR interval of 2DST and M-mode recordings did not differ significantly (P= .35).

The mean values of peak timing are provided in supporting information Table 1. Some remarkable segmental differences were present. SLG was reached significantly later in the basal septal segment compared with the other segments (P < .001) (Fig 3E). The mean delay of this segment was 60 ms (95% confidence interval 37–84 ms). Consequently, SLG occurred after AVCa in 107 of the 139 basal septal segments and in only 74 of the 435 other segments. At the onset of systole, 96% of the analyzed basal septal segments showed a small positive peak strain (SLP), compared with only 35% of the other segments. Time to peak VLA and SrLA was significantly delayed in the lateral wall compared with the septum (P < .001) (Fig 3F). The difference of means for VLA and SrLA was 31 ms (95% confidence interval 20–41 ms) and 25 ms (95% confidence interval 16–34 ms), respectively.


As shown in Tables 1 and 2, the variability of segmental, global, and averaged peak systolic values was low to moderate, whereas diastolic measurements showed a moderate to high variability. SL and SrL exhibited a lower variability than VL and DL. Conversely, the variability of SLP was very high (CV 58–226%). The variability of peak timing was low, except for time to peak VLS, SrLS, and SLP, which showed a moderate to high variability (supporting information Table 1).


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgment
  8. References
  9. Supporting Information

This study describes the technique and demonstrates the feasibility of quantifying LV longitudinal function by 2DST in healthy horses at rest. Global and regional systolic myocardial strain, strain rate, velocity, and displacement could be reliably measured using a modified right parasternal 4-chamber view.

The technique of 2DST is based on tracking the movement of acoustic speckles in the 2D grayscale image from one frame to another. An appropriate frame rate is a prerequisite for adequate 2DST-analysis. A frame rate of 41.1 fps was used, which is within the range of frame rates applied in human medicine (30–90 fps)14,15 and recommended by the software manufacturer (40–70 fps).c Lower frame rates result in large frame-to-frame changes in speckle position and a decreased time resolution. This causes impaired tracking quality and underestimation of peak values, especially at increased HRs.6 In this study, a sufficient frame rate was obtained by limiting the image width to 55°, which did not allow visualizing the entire LV throughout the cardiac cycle. The basal and midwall segments were chosen for analysis as these show the largest longitudinal movement. The frame rate was not increased by reducing the line density of the ultrasound beams, as this results in a decreased lateral resolution. Adequate lateral resolution is important for tracking quality, especially when measuring SL from a right parasternal 4-chamber view. The use of more advanced echocardiographic equipment and transducers may result in an improved lateral resolution, thus enabling the acquisition of higher frame rates at a wider imaging sector.

As 2DST is a new technique, a thorough evaluation of repeatability and reproducibility is crucial before clinical use. All systolic peak values showed a low to moderate between-day and interobserver acquisition and measurement variability and can thus be reliably investigated in future clinical studies. Our reliability data are similar to those reported for 2DST in human medicine, small animals, and horses.8,16–18 However, it should be noted that the comparison of different reliability studies is difficult because of different sample size and statistical methods for assessment of reliability. In our study, diastolic measurements were less reliable as tracking was often visually inaccurate in early diastole because of the fast cardiac movement in this phase. Because of the relatively low frame rate, this early diastolic peak was underestimated by 2DST and early and late diastolic values were approximately equal. In reality, the early diastolic peak is probably much higher than the late diastolic peak in horses at rest, as has already been demonstrated for radial diastolic wall motion velocities by TDI.5

2DST analysis of longitudinal function was feasible in all exams, although 4.3% of segments were excluded because of inadequate tracking quality. Similarly, 5/90 segments were excluded at resting HR in a previous study on equine LV longitudinal function by 2DST.12 The percentage of excluded segments in human medicine ranges from 2.2 to 20% in healthy and diseased study populations.15,19,20 In our study, the basal segments were most frequently excluded. It is important to note that some loops contained artifacts in the mitral annular region. Strict guidelines on the image quality in this region will lead to a lower number of excluded segments. Attention should be paid to an optimal gain and contrast for imaging the myocardium, thereby reducing artefacts. It is particularly important to avoid the presence of drop out because of the ribs or the coronary region. Furthermore, the mitral annular ring has to remain visualized throughout the entire cardiac cycle, also during atrial contraction.

One of the main goals of our study was the quantification of global LV longitudinal function. For this purpose global and averaged peak values were evaluated. Although there was a trend for global SLG to be lower than averaged SLG, the difference was small (mean bias −0.5%) and statistically insignificant. Global SrL values were 5–16% lower compared with the averaged SrL values. Global and averaged SLG and SrLS measurements were equally reliable. However, global values are computed automatically while averaged values need to be calculated manually. Therefore, global SL and SrL values are more attractive to be used in a clinical setting. VL and DL are less applicable as measurements of global LV function because of large segmental differences. Because of the fixed position of the apex in the thorax, the mitral annular ring is pulled down toward the apex during systole. As a result, VL and displacement are largest in segments further away from the apex. The downward motion of the LV base not only results in ejection of blood from the ventricle, but also in passive filling of the atria. The longitudinal base-to-apex gradient is also present in human and small animals.20–22 SL and SrL values did not show this gradient, indicating that all segments shorten almost equally. However, peak SL was significantly lower in the basal septal segment. The basal septal regions also exhibited the lowest contractile indices in the canine and human normal LV.17,23

Our results demonstrated significant segmental differences of peak timing. Although there was a trend of AVCa to occur earlier than AVCMM, the difference was clinically irrelevant (mean bias −10 ms) and statistically insignificant. AVCa was thus considered the end of systole. Postsystolic motion, defined as SLG occurring after AVCa, was present in 31% of segments. This is comparable to human medicine, where postsystolic motion is recorded in about 30% of myocardial segments in healthy people.24 The clinical relevance of the observed postsystolic motion in supposedly healthy horses requires further investigation. However, postsystolic shortening was predominantly observed in the basal septal segment. In the normal human LV, peak strain was also found to be 35–55 ms delayed in the basal septum, which is probably caused by a delayed activation of this segment.25,26 This hypothesis is supported by the frequent presence of early systolic SLP in the basal septal segments in our study. This lengthening of the myocardium arises when a delayed activated segment is passively stretched whereas the other segments start to contract. In the basal septal segment, a small SLP can be considered as normal. In healthy horses, the variability of SLP measurements was very high because of the low peak values. However, the presence of a high SLP in other segments than the basal septal might be useful in a clinical setting as a measure of dyssynchrony.

Surprisingly, a small but significant delay of VLA and SrLA in the LV free wall was found. This might be caused by a delayed activation of the left atrium. Because of the location of the sinoatrial node, the right atrium is depolarized before the left atrium. The septum and right ventricular free wall are stretched at the time of right atrial contraction, whereas stretching of the lateral wall is caused by the subsequent contraction of the left atrium.

The clinical value of LV longitudinal function is extensively documented in human medicine. Several studies have demonstrated that longitudinal function is a sensitive marker for impaired LV function.9–11 In asymptomatic patients with mitral valve disease, myocardial dysfunction can be detected by a decrease in longitudinal shortening. This systolic dysfunction is present before changes in conventional echocardiographic measurements occur. The underlying mechanism is the high vulnerability of the subendocardial longitudinal fibers, whereas the circumferential fibers are less prone to myocardial damage and compensate for the loss of longitudinal function.27 Further studies are required to evaluate if equine LV longitudinal function is also altered by valvular regurgitation. As of yet, the only studies on the use of 2DST in equine cardiology were conducted on healthy adult horses at rest and after exercise.8,12 SLG, SrLS, peak systolic longitudinal displacement (DLS), and time to peak SLG have been used to quantify LV longitudinal function. At resting HRs, peak values, and timing were similar to those reported in our study. Significant alterations were demonstrated during postexercise stress echocardiography. However, it remains to be determined whether 2DST measurements can be used to detect stress-induced hypokinesia, akinesia, or LV dyssynchrony.

The main limitations of our study were inherent to the 2DST technique. An adequate image quality is a prerequisite for accurate tracking. Artefacts such as reverberations and drop out should be avoided during image acquisition, as they have a detrimental effect on the tracking quality. A thorough visual assessment of image quality is recommended before starting offline analysis. A more hidden drawback of 2DST is the extensive smoothing of the curves. This might result in normally appearing curves even when visual tracking quality is poor. For this reason, visual evaluation of tracking quality is indispensable and should be performed before approving any results.

Measurements of longitudinal function were not compared with a gold standard, as there is no such technique available in horses. However, 2DST has been extensively validated in vitro, in animal models, and in human medicine, with sonomicrometry and MRI.19,28,29 Therefore, it can be assumed that the 2DST measurements are closely correlated to the actual deformation of the equine ventricular walls. It should be noted that our results may only be valid for 2DST analyses with the same ultrasound machine, transducer, and offline analysis software and settings.

Our study was performed on 10 horses. A larger study population would have allowed to establish more accurate reference values and to identify influences of body weight, height, breed, sex, and training. This was beyond the scope of this study. Nevertheless, based on our results, preliminary reference values for healthy horses could be formulated.

In conclusion, 2DST was found to be a reliable technique for measuring LV systolic SL, SrL, VL, and DL. These measurements offer new insights into equine ventricular motion and enable a more complete quantification of myocardial function. Further research should be performed in a larger population of healthy and diseased horses to evaluate the value of this technique in horses with cardiac disease.


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgment
  8. References
  9. Supporting Information

aGE Vivid 7 Dimension, GE Healthcare, Horten, Norway

b3S Phased Array Transducer, GE Healthcare

cEchoPAC Software Version7.1.2, GE Healthcare

dSPSS Statistics 17.0, Rel. 17.0.1. 2008, SPSS Inc, Chicago, IL


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgment
  8. References
  9. Supporting Information

The author is a PhD Fellow of the Research Foundation Flanders (FWOVlaanderen).


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgment
  8. References
  9. Supporting Information
  • 1
    Long KJ, Bonagura JD, Darke PG. Standardised imaging technique for guided M-mode and Doppler echocardiography in the horse. Equine Vet J 1992;24:226235.
  • 2
    Patteson MW, Gibbs C, Wotton PR, et al. Echocardiographic measurements of cardiac dimensions and indices of cardiac function in normal adult thoroughbred horses. Equine Vet J 1995;19 (Suppl):1827.
  • 3
    Sepulveda MF, Perkins JD, Bowen IM, et al. Demonstration of regional differences in equine ventricular myocardial velocity in normal 2-year-old thoroughbreds with Doppler tissue imaging. Equine Vet J 2005;37:222226.
  • 4
    Gehlen H, Iversen C, Stadler P. Tissue Doppler echocardiographic examinations at rest and after exercise in horses with atrial fibrillation. Pferdeheilkunde 2009;25:1116.
  • 5
    Schwarzwald CC, Schober KE, Bonagura JD. Methods and reliability of tissue Doppler imaging for assessment of left ventricular radial wall motion in horses. J Vet Intern Med 2009;23:643652.
  • 6
    Teske AJ, De Boeck BW, Melman PG, et al. Echocardiographic quantification of myocardial function using tissue deformation imaging, a guide to image acquisition and analysis using tissue Doppler and speckle tracking. Cardiovasc Ultrasound 2007;5:2745.
  • 7
    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;1:154170.
  • 8
    Schwarzwald CC, Schober KE, Berli AS, et al. Left ventricular radial and circumferential wall motion analysis in horses using strain, strain rate, and displacement by 2D speckle tracking. J Vet Intern Med 2009;23:890900.
  • 9
    Lancellotti P, Cosyns B, Zacharakis D, et al. Importance of left ventricular longitudinal function and functional reserve in patients with degenerative mitral regurgitation: Assessment by two-dimensional speckle tracking. J Am Soc Echocardiogr 2008;21:13311336.
  • 10
    Mizuguchi Y, Oishi Y, Miyoshi H, et al. The functional role of longitudinal, circumferential, and radial myocardial deformation for regulating the early impairment of left ventricular contraction and relaxation in patients with cardiovascular risk factors: A study with two-dimensional strain imaging. J Am Soc Echocardiogr 2008;21:11381144.
  • 11
    Marciniak A, Sutherland GR, Marciniak M, et al. Myocardial deformation abnormalities in patients with aortic regurgitation: A strain rate imaging study. Eur J Echocardiogr 2009;10:112119.
  • 12
    Schefer KD, Bitschnau C, Weishaupt MA, et al. Quantitative analysis of stress echocardiograms in healthy horses with 2-dimensional (2D) echocardiography, anatomical M-mode, tissue Doppler imaging, and 2D speckle tracking. J Vet Intern Med 2010;24:918931.
  • 13
    Schwarzwald CC, Schober KE, Bonagura JD. Methods and reliability of echocardiographic assessment of left atrial size and mechanical function in horses. Am J Vet Res 2007;68:735747.
  • 14
    D'Hooge J, Bijnens B. The principles of ultrasound based motion and deformation estimation. In: SutherlandGR, HatleL, ClausP, eds. Doppler Myocardial Imaging: A Textbook, 1st ed. Hasselt, Belgium: BSWK; 2006:2348.
  • 15
    Suffoletto MS, Dohi K, Cannesson M, et al. Novel speckle-tracking radial strain from routine black-and-white echocardiographic images to quantify dyssynchrony and predict response to cardiac resynchronization therapy. Circulation 2006;113:960968.
  • 16
    Chetboul V, Serres F, Gouni V, et al. Radial strain and strain rate by two-dimensional speckle tracking echocardiography and the tissue velocity based technique in the dog. J Vet Cardiol 2007;9:6981.
  • 17
    Hurlburt HM, Aurigemma GP, Hill JC, et al. Direct ultrasound measurement of longitudinal, circumferential, and radial strain using 2-dimensional strain imaging in normal adults. Echocardiography 2007;24:723731.
  • 18
    Thorstensen A, Dalen H, Amundsen BH, et al. Reproducibility in echocardiographic assessment of the left ventricular global and regional function, the HUNT study. Eur J Echocardiogr 2010;11:149156.
  • 19
    Amundsen BH, Helle-Valle T, Edvardsen T, et al. Noninvasive myocardial strain measurement by speckle tracking echocardiography: Validation against sonomicrometry and tagged magnetic resonance imaging. J Am Coll Cardiol 2006;47:789793.
  • 20
    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;17:10211029.
  • 21
    Chetboul V, Sampedrano CC, Gouni V, et al. Ultrasonographic assessment of regional radial and longitudinal systolic function in healthy awake dogs. J Vet Intern Med 2006;20:885893.
  • 22
    Chetboul V, Sampedrano CC, Tissier R, et al. Quantitative assessment of velocities of the annulus of the left atrioventricular valve and left ventricular free wall in healthy cats by use of two-dimensional color tissue Doppler imaging. Am J Vet Res 2006;67:250258.
  • 23
    Haendchen RV, Wyatt HL, Maurer G, et al. Quantitation of regional cardiac function by two-dimensional echocardiography. I. Patterns of contraction in the normal left ventricle. Circulation 1983;67:12341245.
  • 24
    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;16:415423.
  • 25
    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;27:10871097.
  • 26
    Sengupta PP, Krishnamoorthy VK, Korinek J, et al. Left ventricular form and function revisited: Applied translational science to cardiovascular ultrasound imaging. J Am Soc Echocardiogr 2007;20:539551.
  • 27
    Lee R, Hanekom L, Marwick TH, et al. Prediction of subclinical left ventricular dysfunction with strain rate imaging in patients with asymptomatic severe mitral regurgitation. Am J Cardiol 2004;94:13331337.
  • 28
    Sivesgaard K, Christensen SD, Nygaard H, et al. Speckle tracking ultrasound is independent of insonation angle and gain: An in vitro investigation of agreement with sonomicrometry. J Am Soc Echocardiogr 2009;22:852858.
  • 29
    Korinek J, Wang J, Sengupta PP, et al. Two-dimensional strain—a Doppler-independent ultrasound method for quantitation of regional deformation: Validation in vitro and in vivo. J Am Soc Echocardiogr 2005;18:12471253.

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgment
  8. References
  9. Supporting Information

Table S1. Segmental and averaged peak timing of 2DST measurements of LV longitudinal function in 10 healthy adult trotter horses.

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

Disclaimer: Supporting information is published as submitted and not corrected or checked for scientific content, typographical errors, or functionality. The responsibility for scientific accuracy remains entirely with the authors.

JVIM_663_sm_tables1.doc101KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.