aSchwarzwald CC, Schober KE, Bonagura JD. Echocardiographic characterization of left ventricular radial wall motion in horses using tissue Doppler imaging: Methodology and reliability. J Vet Int Med 2007;21:590.
Methods and Reliability of Tissue Doppler Imaging for Assessment of Left Ventricular Radial Wall Motion in Horses
Article first published online: 25 MAR 2009
Copyright © 2009 by the American College of Veterinary Internal Medicine
Journal of Veterinary Internal Medicine
Volume 23, Issue 3, pages 643–652, May/June 2009
Total views since August 2010: 9
How to Cite
Schwarzwald, C.C., Schober, K.E. and Bonagura, J.D. (2009), Methods and Reliability of Tissue Doppler Imaging for Assessment of Left Ventricular Radial Wall Motion in Horses. Journal of Veterinary Internal Medicine, 23: 643–652. doi: 10.1111/j.1939-1676.2009.0287.x
Previously presented at 25th Annual Forum of the ACVIM, Seattle, WA, June 6–9, 2007.a
- Issue published online: 19 MAY 2009
- Article first published online: 25 MAR 2009
- Submitted August 20, 2008; December 2, 2008; Accepted January 21, 2009.
- Atrial function;
- Index of myocardial performance (Tei index);
- Systolic time intervals;
- Ventricular function;
- Wall motion velocity
Background: Noninvasive assessment of left ventricular (LV) function is incompletely studied in horses.
Objectives: The goals of this study were to investigate the feasibility, techniques, and reliability of tissue Doppler imaging (TDI) for characterization of LV radial wall motion in healthy horses.
Animals: Three Standardbreds, 3 Thoroughbreds; age 8–14 years; body weight 517–606 kg.
Methods: Repeated echocardiographic examinations were performed by 2 observers in unsedated horses using TDI. Test reliability was determined by estimating measurement variability, within-day interobserver variability, and between-day interobserver and intraobserver variability of all echocardiographic variables. Variability was expressed as coefficient of variation (CV) and the absolute value below which the difference between 2 measurements will lie with 95% probability.
Results: Assessment of LV radial wall motion by TDI was feasible in all horses. Measurement variabilities were very low (CV < 5%) to low (CV 5–15%) for most variables. Within-day interobserver variability as well as between-day interobserver and intraobserver variabilities were low to moderate (CV 16–25%) for most variables. All pulsed-wave TDI variables of systolic LV function showed very low to low variability, whereas some of the variables of LV diastolic and LA function showed moderate to high (CV > 25%) variability. Pulsed-wave TDI variables appeared more reliable than color TDI variables.
Conclusions and Clinical Importance: Measurement of TDI indices of LV function is feasible and reliable in adult Standardbred and Thoroughbred horses. The clinical relevance of LV function assessment by TDI remains to be determined.
peak wall motion velocity during late diastole
British standards institution
coefficient of variation
- DT Em
Em deceleration time
- dv/dt Em
Em deceleration slope
early-diastolic transmitral flow velocity
peak wall motion velocity during early diastole
peak wall motion velocity during isovolumic relaxation
index of myocardial performance (Tei index)
isovolumic relaxation time
isovolumic contraction time
left atrial or left atrium
left ventricle or left ventricular
left ventricular ejection time
time to onset of Am
- PW TDI
peak wall motion velocity during ejection
peak wall motion velocity during isovolumic contraction
systolic time interval
time to peak Am
time to onset of Em
tissue Doppler imaging
time to peak Sm
Time to peak S1
Assessment of dimensions and function of the left ventricle (LV) is an important aspect of cardiac evaluation. Echocardiography is the preferred method for clinical assessment of cardiac function because of its widespread availability and its noninvasive nature. However, the functional characteristics of the LV are incompletely studied in horses, and most of the echocardiographic indices used in clinical practice serve to evaluate global systolic LV function, whereas diastolic LV function and regional myocardial function are rarely considered.
Traditionally, a variety of two-dimensional (2D) and M-mode measurements as well as Doppler analyses of intracardiac blood flow have been used to assess systolic LV function in clinical situations and research settings.1–3 LV fractional shortening (FS), assessed in conjunction with measurements of LV dimensions, is the most commonly used index of LV systolic function in horses, and often it is the only one used during routine echocardiography.1,4 Estimation of LV volumes by 2D echocardiography and subsequent calculation of ejection fraction, stroke volume, and cardiac output, is feasible in horses, but is more time consuming, requires precise identification of LV planes over several cardiac cycles, and is based on several geometrical assumptions and approximations.5–7 Volume-based assessment of LV systolic function therefore is not commonly employed in horses.
Systolic LV function also can be assessed by systolic time intervals (STI) measured from M-mode images of aortic and mitral valvular motion or from Doppler tracings of aortic and mitral blood flow, respectively. They include the LV pre-ejection period (PEP), the LV ejection time (LVET), the LV pre-ejection period-to-ejection time ratio (PEP/LVET), and the index of myocardial performance (IMP, also referred to as Tei index).1,8–10 Although STIs may serve as indicators of LV function that may be superior to calculation of FS, they are not always easy to obtain and may only show moderate reliability.1,11,12 Their clinical value has not been well established in horses with cardiovascular disease.
Doppler-derived transmitral flow velocities are commonly used in humans and small animals for assessment of diastolic LV function and filling pressures.1,13 Two recent studies investigated Doppler-derived transmitral flow velocities for assessment of left atrial (LA) mechanical function in adult horses.14,15 Results were in agreement with previous studies,12,16 suggesting that transmitral flow velocity measurements in horses are unreliable and may not be suitable to detect minor changes in LA function and diastolic LV function. Neither the transmitral flow velocities nor other indices of LV diastolic function, such as the isovolumic relaxation time (IVRT),1,13 are commonly used in horses.
Tissue Doppler imaging (TDI) is one of the recent developments in cardiovascular ultrasonography that permits exploitation of many of the newer physical concepts of cardiac function and hemodynamics.3,17,18 It has evolved from a research tool to a clinically applicable method that can be used in addition to conventional echocardiography for assessment of global systolic and diastolic LV function, regional wall motion abnormalities, ventricular synchrony, and filling pressures in humans and small animals.19–27 To date, few reports are available that describe the use of TDI in horses.14,15,28 Therefore, it is unknown whether this new echocardiographic modality will offer any advantage regarding feasibility, reliability, and clinical relevance compared with the conventional 2D, M-mode, and Doppler methods commonly used in horses.
The goal of our study was to demonstrate the feasibility, describe the techniques, and determine the reliability of transthoracic echocardiography for characterization of LV mechanical function in horses by TDI. We hypothesized that TDI methods can be applied to characterize LV radial wall motion in standing, adult horses. Specifically, we aimed to formulate preliminary recommendations for echocardiographic assessment of LV mechanical function in horses by TDI that can be used in future clinical studies to investigate the value of this new echocardiographic method as an adjunct for evaluation of LV systolic and diastolic function.
Material and Methods
Six horses (4 geldings, 2 mares; 3 Standardbreds, 3 Thoroughbreds) aged 9.5 (8–14) years (median, range) and with body weight of 543 (517–606) kg were studied prospectively. All horses were part of the teaching herd of Ohio State University, College of Veterinary Medicine. They were considered healthy based upon physical examination, cardiac auscultation, electrocardiogram, and 2D and M-mode echocardiographic studies. None of the horses was in athletic condition, and none of them received medications during the 2 weeks before entry into the study. The studies were approved by the Institutional Animal Care and Use Committee of Ohio State University.
All studies were conducted in unsedated horses standing in a quiet room, restrained by an experienced handler. Transthoracic echocardiography was performed with a high-end, digital echocardiographb with a phased array transducerc at a frequency of 1.9/4.0 MHz (octave harmonics). A single lead electrocardiogram was recorded simultaneously. Recordings were stored as still frames or cine-loops in digital raw data format for offline analyses.d Three representative, nonconsecutive cardiac cycles were recorded, measured, and subsequently averaged for each variable. Cycles immediately after a sinus pause or 2nd degree atrio-ventricular block were excluded from analyses.
Routine transthoracic 2D, M-mode, and color Doppler echocardiography was performed to assess cardiac structures, valvular competence, chamber dimensions, and LV systolic function, using standard right parasternal long-axis and short-axis views.1,14,29 Transmitral flow velocity profiles were recorded from a left parasternal long-axis view with the pulsed-wave (PW) Doppler cursor positioned between the opened tips of the mitral leaflets.12,14,16 The transducer was positioned as ventrally as possible and angled dorsally to improve alignment with blood flow. No angle correction was used. The peak velocity of early transmitral flow (E wave) was measured for subsequent calculation of E to peak early-diastolic wall motion velocity (Em) ratio (E/Em; Appendix 1).
The main attention then was directed to assessment of LV wall motion by TDI. The LV was imaged in tissue velocity imaging modeb using a right parasternal short-axis view at the level of the chordae tendineae. For the pulsed-wave TDI (PW TDI) recordings, a sample volume 5.9 mm in width was used. The sample volume was placed on the LV free wall so that it covered the subendocardial region during diastole and stayed on the myocardium throughout the cardiac cycle. The velocity scale was set to −30 to +20 cm/s. The simultaneous 2D image was frozen during the PW TDI recordings to improve image quality (Fig 1A). For measurements, the sweep speed was adjusted to display a single cardiac cycle. The outer edge of the strongest echo was measured at standard gain settings.
For the color TDI (cTDI) recordings, the imaging depth and sector width, respectively, were adjusted to achieve a frame rate of at least 120 frames/s while the region of interest was maximized to cover the entire imaging sector and the velocity scale was set to −35 to +35 cm/s. Analyses of the cTDI recordings were performed by the Q-Analysis function in trace mode.d The image was frozen and the cursor placed at the onset of the QRS complex shown on the ECG. The sweep speed was adjusted so that 1 cardiac cycle filled the entire screen. A 20 × 30 mm sample volume was placed on the LV free wall as shown in Figure 1B, so that it covered the entire LV wall thickness during diastole. The anchor function was used to allow tracking of myocardial motion throughout the cardiac cycle. A standard 30 ms smoothing filter was applied.
The echocardiographic variables used in this study included several measured and calculated variables that were derived from PW TDI and cTDI mode, respectively (Appendix 1). The variables were grouped into 9 variables of LV systolic function, 1 index of combined LV systolic and diastolic function, 8 variables of LV diastolic function, and 5 variables of LA function. Our pilot studies had shown that calculation of transmural velocity gradients, radial strain, and radial strain rate by cTDI and Q-Analysis was prone to error and thus unreliable (data not shown). Therefore, this study was limited to evaluation of radial velocity variables. For some of the variables, triplicate measurement was not possible because of inability to clearly identify the respective velocity waves on the available cycles. The affected data points were reported as missing (Tables 1 and 2).
|Measurement Variability||Within-Day Variability||Between-Day Variability|
|Variables||Units||Missing Data Pointsa||Mean ± SDb||CV (%)||BSI||CV (%)||BSI||CV (%)||BSI||CV (%)||BSI||CV (%)||BSI|
|Variables of systolic LV function|
|S1||cm/s||1||9.77 ± 1.30||2.1||0.573||11.1||3.25||10.4||2.67||11.3||3.01||6.3||1.68|
|Sm||cm/s||0||12.11 ± 1.68||0.7||0.239||10.1||3.65||4.8||1.49||10.5||3.47||11.7||3.87|
|tS1||ms||1||59.6 ± 12.7||3.6||6.00||14.3||25.9||4.6||7.29||14.3||22.6||14.0||22.3|
|tSm||ms||0||213.4 ± 35.7||4.3||25.6||6.4||39.1||6.4||35.8||5.2||29.6||5.6||32.1|
|PEP||ms||0||126.9 ± 25.3||2.4||8.40||6.5||23.4||11.0||38.2||6.2||21.4||6.7||23.0|
|IVCT||ms||1||90.6 ± 24.9||3.4||8.53||7.7||18.7||13.4||35.5||12.4||32.4||12.1||29.7|
|ET||ms||0||404.0 ± 18.9||0.9||9.47||1.5||16.8||3.6||39.9||6.0||67.1||2.9||31.6|
|PEP/ET||—||0||0.32 ± 0.07||3.1||0.0266||7.9||0.0705||9.9||0.0848||6.9||0.0589||8.8||0.0755|
|IVCT/ET||—||1||0.23 ± 0.07||3.8||0.0239||8.4||0.0508||11.9||0.0782||10.8||0.0699||13.4||0.0827|
|Index of systolic and diastolic LV function|
|IMP||—||1||0.35 ± 0.09||4.5||0.0438||7.0||0.0689||6.9||0.0707||8.7||0.0868||13.7||0.134|
|Variables of diastolic LV function|
|E1||cm/s||0||6.79 ± 1.31||4.3||0.811||17.4||3.66||13.5||2.62||8.5||1.60||10.5||2.07|
|Em||cm/s||0||28.66 ± 2.54||2.7||2.20||5.3||4.19||9.1||7.26||5.3||4.25||9.1||7.27|
|Em/Am||—||0||3.03 ± 0.75||2.7||0.229||15.0||1.15||16.7||1.49||18.3||1.60||11.8||0.975|
|E/Em||—||0||2.04 ± 0.45||2.0||0.113||6.3||0.366||18.3||1.047||4.6||0.259||17.2||0.995|
|dv/dt Emc||cm/s2||0||652.6 ± 111.4||11.5||215||32.4||479||23.0||375||21.6||387||24.4||419|
|DT Emc||ms||0||48.1 ± 9.2||12.0||15.6||36.3||62.6||16.2||23.7||18.8||25.6||23.9||33.5|
|IVRT||ms||0||48.4 ± 17.2||6.9||9.55||17.0||25.4||17.2||23.7||10.9||14.8||16.1||22.2|
|tEm||ms||0||579.6 ± 21.8||0.5||8.05||2.7||44.3||5.0||80.9||4.2||67.9||3.9||62.9|
|Variables of active LA function|
|Am||cm/s||0||9.97 ± 2.35||2.3||0.626||13.6||4.1||12.2||3.29||12.5||3.44||7.7||2.22|
|tAm||ms||0||209.7 ± 25.2||0.8||4.67||2.4||14.2||8.2||43.2||16.3||89.9||12.6||67.7|
|pAm||ms||0||149.4 ± 18.5||3.4||13.7||2.9||12.3||13.9||51.1||18.1||70.9||14.1||52.5|
|dAm||ms||0||132.3 ± 14.0||2.6||9.52||3.4||12.6||7.8||28.2||5.5||19.7||7.1||25.9|
|pAm/dAm||—||0||1.15 ± 0.21||5.3||0.165||4.7||0.150||18.7||0.536||16.7||0.513||16.4||0.477|
|Measurement Variability||Within-Day Variability||Between-Day Variability|
|Variables||Units||Missing Data Pointsa||Mean ± SDb||CV (%)||BSI||CV (%)||BSI||CV (%)||BSI||CV (%)||BSI||CV (%)||BSI|
|Variables of systolic LV function|
|S1d||cm/s||2||5.27 ± 1.42||4.8||0.726||8.8||1.28||9.6||1.34||16.2||2.35||14.8||2.14|
|Sm||cm/s||0||8.82 ± 1.59||1.3||0.313||1.9||0.458||14.9||3.56||15.8||3.85||9.8||2.58|
|tS1d||ms||2||59.1 ± 15.8||10.7||17.0||25.1||48.3||14.7||25.5||14.4||25.4||9.8||16.8|
|tSm||ms||0||223.3 ± 35.6||3.9||23.7||6.1||38.9||8.1||50.0||5.8||36.5||7.5||45.2|
|PEP||ms||0||114.2 ± 25.7||6.5||19.9||14.3||49.8||7.1||23.0||6.7||21.4||7.7||24.1|
|IVCT d||ms||6||77.2 ± 17.5||6.4||13.6||9.7||21.3||9.6||22.7||6.6||15.3||9.7||21.6|
|ET||ms||0||419.6 ± 16.7||1.4||16.3||1.7||20.1||5.3||59.9||6.9||78.6||3.0||33.7|
|PEP/ET||—||0||0.28 ± 0.07||7.3||0.0535||13.9||0.115||4.7||0.0374||12.5||0.0968||12.2||0.0923|
|IVCT/ET d||—||6||0.18 ± 0.05||6.2||0.0317||10.6||0.0549||9.6||0.0553||18.2||0.101||17.0||0.0915|
|Variables of systolic and diastolic LV function|
|IMPd||—||12||0.35 ± 0.05||4.6||0.0402||4.3||0.0387||n/dc||n/dc||13.9||0.130||17.8||0.175|
|Variables of diastolic LV function|
|E1d||cm/s||6||4.22 ± 0.83||9.8||1.05||17.3||2.02||34.6||3.90||22.7||2.66||33.6||4.17|
|Em||cm/s||0||21.38 ± 4.13||3.8||2.26||5.9||3.55||14.5||8.46||14.4||8.33||16.9||9.97|
|Em/Am||—||0||3.61 ± 1.21||6.3||0.644||4.6||0.457||16.9||1.54||18.6||1.79||20.8||1.83|
|E/Em||—||0||2.88 ± 0.97||2.0||0.161||6.6||0.529||21.0||1.667||19.2||1.562||16.9||1.371|
|dv/dt Emd||cm/s2||0||298.7 ± 112.5||8.2||70.3||22.0||182||26.5||203||25.9||200||36.1||272|
|DT Emd||ms||0||84.4 ± 28.3||5.0||11.4||25.7||57.6||17.1||39.2||21.9||51.9||15.2||38.1|
|IVRT d||ms||5||56.4 ± 18.4||10.3||15.3||10.9||16.4||19.7||32.3||7.1||11.2||21.0||32.2|
|tEm||ms||1||588.5 ± 14.8||2.0||33.1||2.9||48.6||5.9||93.3||4.8||76.5||2.5||40.1|
|Variables of active LA function|
|Am||cm/s||0||6.45 ± 1.61||3.0||0.540||2.1||0.374||20.6||3.89||14.2||2.55||9.9||2.01|
|tAm||ms||0||216.0 ± 26.7||2.9||17.3||3.2||19.4||9.6||53.5||10.5||60.9||9.3||51.8|
|pAm||ms||0||122.0 ± 16.5||5.1||17.0||9.0||31.9||18.1||54.0||17.2||56.2||16.0||48.6|
|dAm||ms||0||158.4 ± 16.8||3.7||16.4||8.5||37.0||8.2||34.6||8.2||35.8||5.9||25.3|
|pAm/dAm||—||0||0.79 ± 0.10||7.7||0.167||16.1||0.375||23.7||0.477||20.4||0.443||12.6||0.256|
The minimal resolution of measurements, hence the smallest possible increments of measurement given by the software, varied slightly depending on velocity scale (which was variable for cTDI recordings because of the auto-scaling feature of the Q-Analysis software) and sweep speed (which was chosen individually to display 1 single cardiac cycle). In summary, the smallest increments of measurement were 1 cm/s (velocity) and 2–4 ms (time intervals) for PW TDI recordings; 0.04–0.06 cm/s and 3–4 ms for cTDI recordings; and 1 cm/s for PW Doppler recordings of transmitral flow velocity profiles.
Reliability of Echocardiographic Variables
All horses underwent repeated echocardiographic examinations by 2 experienced examiners, according to previously determined imaging guidelines. One echocardiographer (CCS) examined each horse twice at an interval of 2 days. On 1 occasion, a 2nd, independent echocardiographer (KES) examined each horse immediately before (3 horses) or after (3 horses) the other echocardiographer. All recordings were labeled with random codes, allowing subsequent offline measurements in a blinded fashion.
The intraobserver measurement variability was determined by a single observer (CCS) measuring the same 6 studies (1 study of each horse) repeatedly on 2 different days, thereby averaging the same set of 3 cardiac cycles for each variable. For determination of the interobserver measurement variability, a 2nd observer (KES) measured the same cardiac cycles on the same 6 studies, independently of the 1st observer. For determination of the within-day interobserver variability, 1 observer (CCS) measured the 2 studies of each horse that were recorded consecutively on the same day by the 2 observers. The between-day intraobserver variability was determined by 1 blinded observer (CCS) measuring each horse's 2 studies that were recorded by CCS on different days. The between-day interobserver variability was determined by 1 blinded observer (CCS) measuring 2 studies of each horse that were recorded by the 2 observers 2 days apart; the studies were chosen so that 3 horses were examined by CSS first and 3 horses were examined by KES first. All measurements were performed with the stored recordings in random order and with the observers blinded to signalment and previous measurements.
All statistical and graphical analyses were performed with standard computer software.e,f,g Test reliability was quantified by the within-subject variance for repeated measurements (residual mean square) determined by 1-way analysis of variance with the horses as groups.30 The within-subject standard deviation (sw) was calculated as the square root of the residual mean square. Measurement variability and recording variability were reported in 2 ways: (1) The within-subject coefficient of variation (CV) expressed as a percent value was calculated as CV =sw/mean × 100 in order to compare the reliability of the various variables in this study.30 The degree of variability was arbitrarily defined as follows: CV < 5%, very low variability; 5–15%, low variability; 16–25%, moderate variability; > 25%, high variability. (2) In addition to the CV, the absolute value below which the difference between 2 measurements will lie with 95% probability was estimated following the British Standards Institution (BSI) recommendations: BSI = 1.96 ×√2 ×sw= 2.77 ×sw.30 The BSI was reported to provide a clinically applicable measure of variability, hence an absolute value that allows comparison with measured changes in echocardiographic variables on a case-by-case basis.
Summary statistics (mean ± SD) for each variable were calculated based on the 1st study of each horse (n = 6) and were reported for comparison. For comparison of the variability of the PW TDI variables and the cTDI variables, respectively, the data of the various categories of variability (eg, measurement variability, within-day variability, between-day variability) were pooled and the CVs of each variable were compared by a Wilcoxon's signed rank test. The level of significance was defined as α= 0.05.
Preliminary recommendations for the use of the TDI variables in future studies were formulated based on reliability data and the number of missing data points. Variables with high variability (ie, CV > 25%) in any of the categories of variability and variables with more than 1 missing data point were not recommended for further use.
Results of the routine 2D and M-mode echocardiography (median, minimum to maximum) were as follows: heart rate, 37 beats/min (34–42); maximum LA diameter, 12.4 cm (10.9–13.8); maximum LA area in long-axis view, 97 cm2 (70–103); maximum LA area in short-axis view, 112 cm2 (91–136); interventricular septal thickness in diastole, 3.3 cm (2.7–3.4); LV inner diameter in diastole, 11.1 cm (10.3–12.7); LV free wall thickness in diastole, 2.5 cm (2.3–3.0); LV fractional shortening, 38% (31–52); diameter of the aortic sinus at end-diastole, 7.1 cm (6.6–8.2); and diameter of the pulmonary artery sinus at end-diastole, 6.4 cm (6.2–6.8).
Echocardiographic assessment LV radial wall motion by TDI was feasible in all horses using right-parasternal short-axis views. The frame rate during cTDI recordings ranged from 146.4 to 160.8 frames/s. Recordings of adequate quality could be obtained in all horses with both PW TDI and cTDI methodology. The LV radial wall motion velocity profile during systole was characterized by an early-systolic positive wave during isovolumic contraction (S1), followed by a 2nd positive wave during ventricular ejection (Sm) (Fig 1). The LV radial wall motion velocity profile during diastole was characterized by an early-diastolic negative wave during isovolumic relaxation (E1), a 2nd negative wave during rapid ventricular filling (Em), and a late-diastolic wave during atrial contraction (Am). The Em wave and the Am wave were followed by small positive velocity waves, reflecting passive recoil of the LV. This wall motion pattern appeared very consistently in all horses. Generally, identification of the different wall motion velocity waves was easier in the PW TDI recordings, whereas the beginning and end of the velocity waves often were less distinct in cTDI recordings. This was particularly true for the short-lived isovolumic velocity waves S1 and E1, respectively. Consequently, more missing data points were present among the cTDI measurements compared with the PW TDI measurements (Tables 1 and 2).
Reliability data of all echocardiographic variables are summarized in Tables 1 and 2. Briefly, intraobserver measurement variability was very low to low for all variables. Interobserver measurement variability was very low to low for the majority of variables, with some exceptions showing moderate to high variability. Within-day interobserver variability as well as between-day intraobserver and between-day interobserver variabilities, respectively, were low to moderate for most variables, with few exceptions in the very low and in the high range.
All PW TDI-derived variables of systolic LV function showed very low to low variability, whereas some of the variables of LV diastolic function and LA function showed moderate to high variability. Overall, based on the pooled data of all variables, the PW TDI variables were more reliable than the cTDI variables (P= .0007).
Results of this investigation indicated that TDI analyses are both feasible and reliable for assessment of radial motion of the LV free wall from short-axis images recorded at the chordal level in standing, unsedated, adult Standardbred and Thoroughbred horses. Most of the TDI variables used in this study had sufficient test-retest reliability to justify further investigations into their use for the objective quantification of systolic and diastolic LV function in horses.
In people, TDI-based markers of decreased long-axis contraction have been shown to unmask subclinical systolic LV dysfunction in asymptomatic patients with aortic and mitral regurgitation.31–33 Furthermore, TDI has been used to detect systolic LV dysfunction in patients with diastolic heart failure,34 to uncover asymptomatic patients with myocardial disease,35–37 and to discriminate between hypertrophic cardiomyopathy and athlete's heart.37,38 Similarly, TDI was able to detect early asymptomatic myocardial dysfunction during the preclinical phase of cardiomyopathy associated with muscular dystrophy in dogs.22,39 Whether TDI variables offer similar clinical value in horses and whether they provide any advantages over conventional echocardiographic indices for assessment of LV systolic function cannot be judged at this stage of validation and will have to be investigated in future studies. However, this investigation on the feasibility and reliability of TDI in horses is an important 1st step to the introduction of this new technology in clinical practice.
Details on calculation and interpretation of indices of reliability have been discussed elsewhere.14 The results of this study indicated that all variables had very low to low intraobserver measurement variability, indicating that image quality and measurement guidelines were sufficient for all variables. However, the interobserver measurement variability was higher and ranged from very low to high. The within-day, interobserver variability ranged between very low and moderate for PW TDI variables and between very low and high for cTDI variables. The overall higher variability compared with the intraobserver measurement variability was primarily attributed to the error introduced by more than 1 operator performing the examination, and an additional effect of short-term biological variability (ie, biological changes occurring within a few hours of examination) also may have influenced the results. These findings emphasize the importance of well defined imaging guidelines and adequate operator training that should allow independent observers to record standardized images of adequate quality and to perform reliable measurements. The results suggest that a single operator should perform sequential recordings and measurements in the same patient to minimize test-retest variability.
The between-day intraobserver variability and the between-day interobserver variability, respectively, were determined to assess the day-to-day variability of measurements, including biological, recording, and measurement variability. Both ranged between very low and moderate for PW TDI variables and between very low and high for cTDI variables. These 2 measures of test reliability may be most important for clinical applications. In general, true alterations in echocardiographic variables caused by disease or interventions in an individual patient would have to be larger than the potential changes simply caused by day-to-day variability (represented by the BSI value).
Theoretically, cTDI recordings offer certain advantages over PW TDI recordings. For instance, cTDI allows tracking of the region of interest by anchoring of the sample volume to the myocardial wall, simultaneous investigation of regional motion of multiple wall segments, and calculation of transmural velocity gradients as well as strain and strain rate.18 However, our pilot studies showed that radial transmural velocity gradients, radial strain, and radial strain rate cannot be reliably measured in horses (data not shown). Furthermore, cTDI analyses are more time-consuming, require specific analysis software, and are influenced by a variety of filter settings. Also, the high frame rate that is necessary for accurate cTDI recordings (ie, > 120 frames/s) can only be achieved by limiting the imaging depth and sector width to an extent that the LV cannot be imaged in its entirety. The results of this study indicated that cTDI-based indices of LV function had an overall higher variability compared with PW TDI-based indices. However, direct comparison of reliability data may be hampered by fundamental differences in methodology between PW TDI and cTDI. First, the results of the cTDI analyses represented average velocities within the sample volume, whereas the results of the PW TDI analyses represented maximum velocities within the sample volume. Secondly, the sample volume for cTDI recordings was chosen to cover the entire LV wall thickness at end-diastole and was anchored to the myocardium so that it stayed on the region of interest during the entire cardiac cycle. Conversely, for PW TDI analyses, the sample volume was smaller, did not cover the entire myocardium, and was in a fixed position relative to the transducer, so that the myocardium was moving relative to the sample volume during the cardiac cycle. Moreover, the cTDI sample volume was placed during offline analyses, whereas the PW TDI sample volume was placed during recording. Therefore, the measurement error may have been increased for cTDI indices compared with PW TDI indices because of expected variations in sample volume placement during offline analyses. Conversely, within-day interobserver and between-day interobserver variabilities may be decreased for cTDI indices because cTDI sample placement was done by a single observer who performed all offline analyses whereas PW TDI sample placement was done by 2 different observers recording the echocardiograms. Nonetheless, in addition to the apparent differences in reliability, the results of this study also showed a higher number of missing data points for cTDI indices because of difficulties in identifying the beginning and the end of some of short-lived wall motion velocity waves. Therefore, considering all the differences between PW TDI and cTDI, we believe that cTDI recordings do not provide any substantial advantages over PW TDI recordings for analysis of radial LV wall motion in horses.
Overall, the majority of TDI variables of LV systolic function were easy to measure and proved to be reliable in the present study. Eight of the 10 indices of LV systolic function investigated in this study (including the IMP) were time intervals. Systolic time intervals measured by M-mode recordings or by Doppler recordings of intracardiac blood flow are considered simple indices of global ventricular function that are independent of ventricular shape and geometry.1,8–10 They are sensitive to a variety of disease processes and pharmacologic interventions, but are variably affected by heart rate and loading conditions.1,8 Although the use of systolic time intervals has been described in horses,9,11,12,40 they are not commonly used during routine echocardiography in this species. One inherent problem when measuring STIs by conventional echocardiographic methods is the difficulty in accurately identifying the opening and closing points of the valves by M-mode and the onset and cessation of the ventricular inflow and outflow signals by spectral Doppler, respectively. Calculation of the IMP further requires measurements using nonsimultaneous echocardiographic recordings from different imaging planes, adding error because of changes in cardiac cycle length that may occur during data acquisition.41,42 Use of TDI to measure STIs and calculate IMP has been described in other species.41–44 Results of the present study indicate that STIs and IMP can be reliably measured based on TDI recordings of LV radial wall motion. TDI allows analysis of systolic and diastolic wall motion velocity patterns from the same cardiac cycle by a single recording in 1 imaging plane. Moreover, at least in PW TDI mode, it provides well-defined, distinct wall motion velocity waves that allow unambiguous measurement of time intervals. Therefore, data collection and analysis are considerably facilitated compared with conventional methods to determine STIs. Nonetheless, when STIs are derived from TDI analysis of a single LV wall segment, as was done in this study, they may be influenced by local wall motion abnormalities and therefore may not represent true global LV function. The clinical relevance of this limitation cannot be predicted based on this study, but needs to be considered when TDI-based STIs are used in future investigations.
Although assessment of systolic LV function is one of the main goals of echocardiography in horses, diastolic LV function is poorly investigated and often is not critically assessed during routine echocardiography. In people and small animals, TDI has been shown to be particularly useful in conjunction with transmitral flow patterns for detection of impaired LV relaxation, classification of diastolic LV dysfunction, and estimation of LV filling pressures.13,23,26 Similarly, several diastolic TDI variables were investigated in this study as potential indicators of diastolic LV function and filling pressures. Although the results suggest that variables of LV diastolic function using TDI may be somewhat less reliable than those of LV systolic function, some of the investigated indices still may prove clinically useful and may add to the understanding of diastolic ventricular function in horses. In fact, among the PW TDI variables of LV diastole, all but the variables of deceleration of early diastolic blood flow (dv/dt Em, DT Em) were characterized by moderate or lower variability and therefore warrant further investigation.
Assessment of diastolic cardiac performance should not be limited to LV relaxation and early filling, but also should include assessment of atrial mechanical function. Unlike variables of LA function obtained by echocardiographic examination of the left atrium itself,14,15 the variables of LA function investigated in this study represent indirect measures of LA function, because they are based on LV wall motion analysis. They are likely influenced by a variety of factors, including loading conditions, inotropic state, ventricular relaxation and compliance, heart rate, and intra-atrial as well as atrio-ventricular conduction.14,45 Nevertheless, compared with previously reported indices of LA function obtained by TDI-analyses of the LA free wall in horses,14 the reliability of variables of LA function investigated in the present study appeared higher. This may be attributed to the fact that TDI variables of LA function investigated in this study were easier to record and to measure compared those used in the previous study. Nevertheless, it is currently unknown if TDI variables of LA function are of clinical value and if TDI offers any advantage, or at least incremental value, over 2D echocardiography14,15 for assessment of LA function.
In addition to the restrictions already mentioned, there are other limitations of this study that require discussion. First, because of differences in patient population, equipment, machine settings, observer experience, and image quality, the data of our study will not be directly applicable to all clinical situations. A second potential limitation concerns the echocardiographic short-axis views used in this study. In humans, TDI usually is applied to study long-axis motion from LV apical views. Most commonly, longitudinal motion of the mitral annulus is used to study global systolic and diastolic function.19,46 However, both longitudinal motion and radial motion have been studied in other species and both were considered diagnostically valuable by some authors.20,22 In cats, TDI analysis of radial motion of the LV free wall was considered technically easier and seemed to be more reliable compared with TDI analysis of mitral annular longitudinal motion in 1 study.20 In horses, true apical views are impossible to obtain and adequate alignment of the Doppler beam with longitudinal mitral annular motion cannot be achieved. Therefore, we limited our investigations to radial wall motion of the LV free wall that can be assessed by standard short-axis views, allowing optimal alignment with wall motion. This approach resulted in the restriction of TDI analyses to 1 single myocardial segment. Specifically, we investigated motion of a segment of the LV free wall with the intention to use the derived variables as a measure of global LV function. This procedure is similar to the calculation of the LV fractional shortening derived from conventional M-mode images or to the analysis of longitudinal motion of the mitral annulus to study global systolic and diastolic LV function, as commonly performed in humans and small animals.19,46 It remains to be determined if TDI analyses of radial motion of the LV free wall indeed will be able to represent global LV function, and if this technique will be sufficiently sensitive and specific to detect alterations in LV function.
Independent of methodical approach and study design, TDI itself has its own limitations. Specifically, TDI only provides one-dimensional information on ventricular wall motion, ignoring the fact that LV function is determined by complex, three-dimensional motion patterns.3,17,18 Furthermore, the insonation angle may vary during the cardiac cycle, velocity may reflect gross translation rather than actual local contraction, and akinetic segments may show motion caused by tethering.3,18 Therefore, it is important to remember that TDI data provide limited information on wall motion and that velocity data obtained from different views or from different locations should not be used interchangeably. Finally, TDI variables, including velocity indices and time intervals, may be influenced by alterations in heart rate and loading conditions.1,8 Although the present study design did not allow testing rate and load dependency of TDI variables, these properties could be of great importance in clinical applications. Nonetheless, acknowledging all of the limitations related to study design and tissue Doppler methodology, we believe that TDI still may add to the understanding of LV mechnical function and aid in the diagnosis of global LV systolic and diastolic dysfunction in horses.
In conclusion, we were able to show that radial motion of the LV free wall can be reliably characterized by TDI analysis in a right-parasternal short-axis view at the level of the chordae tendineae. We suggest that TDI-based indices may be useful for assessment of LV and LA performance in horses. Our imaging guidelines and recommendations should be regarded as preliminary and need validation in clinical practice. Additional studies will be required to establish reference values, considering potential differences related to breed, age, sex, body weight, and athletic condition. Finally, the clinical value of these variables to assess disease-related alterations in LV and LA function, and their relation to severity of disease, exercise capacity, and prognosis need to be determined.
a Schwarzwald CC, Schober KE, Bonagura JD: Echocardiographic characterization of left ventricular radial wall motion in horses using tissue Doppler imaging: Methodology and reliability. J Vet Int Med 2007; 21:590 (Abstract).
bGE Vivid 7, BTO4, GE Medical Systems, Milwaukee, WI
cM3S phased array transducer, GE Medical Systems
dEchoPAC Software v3.1.3, GE Medical Systems
eMicrosoft Office Excel 2003, Microsoft Corporation, Redmond, WA
fSigmaStat v3.01, SPSS Inc, Chicago, IL
gGraphPad Prism v5.00 for Windows, GraphPad Software, San Diego, CA
- 1Evaluation of size, function, and hemodynamics. In: BoonJA, ed. Manual of Veterinary Echocardiography. Baltimore, MD: Williams & Wilkins; 1998:151–260.
- 2Left and right ventricular systolic function. In: OttoCM, ed. Textbook of Clinical Echocardiography, 3rd ed. Philadelphia, PA: Elsevier Saunders; 2004:131–165.
- 4Echocardiographic measurements of cardiac dimensions and indices of cardiac function in normal adult thoroughbred horses. Equine Vet J 1995;19 (Suppl):18–27., , , et al.
- 5Accuracy of formulae for calculating left ventricular volumes of the equine heart. Equine Vet J 1990;9 (Suppl):53–56.,
- 11Measurement of cardiac function by transthoracic echocardiography: Day to day variability and repeatability in normal Thoroughbred horses. Equine Vet J 1998;30:117–122.,
- 12Pulsed wave Doppler echocardiography in normal horses. Equine Vet J 1995;19 (Suppl):38–46.,
- 13Ventricular diastolic filling and function. In: OttoCM, ed. Textbook of Clinical Echocardiography, 3rd ed. Philadelphia, PA: Elsevier Saunders; 2004:166–195.
- 18http://folk.ntnu.no/stoylen/strainrate/. Accessed July 2008.Strain rate imaging: Cardiac deformation imaging by ultrasound / echocardiography—Tissue Doppler and Speckle tracking. Trondheim, Norway. Available at:
- 29Cardiovascular ultrasonography. In: ReefVB, ed. Equine Diagnostic Ultrasound. Philadelphia, PA: WB Saunders; 1998:215–272.
- 30Clinical measurement. In: BlandM, ed. An Introduction to Medical Statistics, 3rd ed. Oxford, UK: Oxford University Press; 2000:268–293.
- 36Evidence of impaired left ventricular systolic function by Doppler myocardial imaging in patients with systemic amyloidosis and no evidence of cardiac involvement by standard two-dimensional and Doppler echocardiography. Am J Cardiol 2008;101:1039–1045., , , et al.
- 40Effects of sedation with detomidine hydrochloride on echocardiographic measurements of cardiac dimensions and indices of cardiac function in horses. Equine Vet J 1995;19 (Suppl):33–37., , , et al.
- 46Evaluation of systolic and diastolic function of the left ventricle. In: FeigenbaumH, ArmstrongWF, RyanT, eds. Feigenbaum's Echocardiography, 6th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2005:138–180., ,
|Variables of systolic LV function|
|Systolic wall motion velocities|
|S1||Isovolumic contraction velocity||Peak positive wall motion velocity during isovolumic contraction|
|Sm||Ejection velocity||Peak positive wall motion velocity during ejection|
|Systolic time intervals|
|tS1||Time to peak S1||Time from the onset of the QRS complex to peak S1|
|tSm||Time to peak Sm||Time from the onset of the QRS complex to peak Sm|
|PEP||Pre-ejection period||Time from the onset of the QRS complex to the onset of the Sm wave*|
|IVCT||Isovolumic contraction time||Time from the onset of the S1 wave to the onset of the Sm wave*|
|ET||Ejection time (Sm duration)||Time from the onset to the end of the Sm wave*|
|Index of systolic and diastolic LV function|
|IMP||Index of myocardial performance|
|(Tei index)||IMP = (IVCT + IVRT) / ET|
|Variables of diastolic LV function|
|Diastolic wall motion velocities|
|E1||Isovolumic relaxation velocity||Peak negative wall motion velocity during isovolumic relaxtion|
|Em||Early-diastolic velocity||Peak negative wall motion velocity during the phase of rapid ventricular filling|
|E/Em||E-to-Em ratio||Ratio of early-diastolic transmitral flow velocity (E) over Em|
|dv/dt Em||Em deceleration slope||Slope of the Em deceleration|
|DT Em||Em deceleration time||Time from peak to the end Em*|
|Diastolic time intervals|
|IVRT||Isovolumic relaxation time||Time from the onset of the E1 wave to the onset of the Em wave*|
|tEm||Time to onset of Em||Time from the onset of the QRS complex to the onset of the Em wave*|
|Variables of active LA function|
|Am||Late-diastolic velocity||Peak negative wall motion velocity at the time of atrial contraction|
|tAm||Time to peak Am||Time from the onset of the P wave to peak Am|
|pAm||Time to onset of Am||Time from the onset of the P wave to the onset of the Am wave*|
|dAm||Am duration||Time from the onset to the end of the Am wave*|