Comparison of 2 Myocardial Velocity Gradient Assessment Methods During Dobutamine Infusion with Doppler Myocardial Imaging

Comparison of 2 Myocardial Velocity Gradient Assessment Methods During Dobutamine Infusion with Doppler Myocardial Imaging Denis Pellerin, MD, FESC, Alain Berdeaux, MD, Laurent Cohen, PhD, Jean F. Giudicelli, MD, Serge Witchitz, MD, and Colette Veyrat, MD, Le Kremlin-Bicetre and Paris, France

Myocardial velocity gradient (MVG) has been shown to be the best quantitative parameter for the detection of ischemic myocardium during dobutamine infusion with the use of Doppler myocardial imaging. MVG has been previously assessed by velocity measurements across the thickness of the myocardium at the time of visually selected maximal color brightness (thickness-velocity plot method). We hypothesized that MVG could be assessed by velocity measurements throughout the cardiac cycle in the subendocardium parallel to the endocardial boundary to the left ventricular cavity and in the subepicardium parallel to the epicardial boundary (time-velocity plot method). This study was designed to compare MVG obtained from the thickness-velocity plot method and from the time-velocity plot method in quantifying dobutamine-induced changes in myocardial wall motion in 8 phases of the cardiac cycle on color Mmode Doppler myocardial imaging recordings of the left ventricular posterior wall performed in 8 conscious dogs at baseline and at steady state during dobutamine infusion (10 µg/kg per minute). For both methods, MVG was considered present if its mean value was significantly different from zero and if

Doppler myocardial imaging (DMI) provides quantitative assessment of systolic and diastolic velocities within the myocardium. Myocardial velocity gradient (MVG) has been used as a quantitative parameter to From the Department of Cardiology and Department of Pharmacology, University Hospital Bicetre, Paris-Sud University Medical School, Le Kremlin-Bicetre, France; the Department of Image Processing, Ceremade, Paris-Dauphine University, Paris, France; and CNRS/Inserm U141, University Hospital Lariboisiere, Paris, France. Reprint requests: Dr Denis Pellerin, Service de Cardiologie, Hopital Universitaire de Bicetre, 78, rue du General Leclerc, 94270 Le Kremlin-Bicetre, France. Copyright © 1999 by the American Society of Echocardiography. 0894-7317/99/$8.00 + 0 27/1/94835

22

endocardial and epicardial velocities were significantly different. There was close agreement between the 2 methods. MVG was present during the preejection period, systole, rapid ventricular filling, and atrial contraction. Dobutamine induced a significant increase in MVG during the preejection period (from 2.64 ± 0.83 to 4.05 ± 0.81 seconds -1), systole (from 2.14 ± 0.59 to 6.08 ± 2.20 seconds-1 in early systole, from 1.90 ± 1.06 to 5.31 ± 2.95 seconds-1 in mid systole, from 1.37 ± 0.57 to 2.44 ± 0.53 seconds-1 in end systole), and rapid ventricular filling (from 3.06 ± 1.12 to 7.82 ± 2.58 seconds-1), related to a greater rise in endocardial than in epicardial velocities. The time-velocity plot method showed that ejection and diastole were 11% and 28% decreased during dobutamine infusion, respectively, as heart rate was 31% increased. Thus according to our quantitative criteria, both MVG assessment procedures enabled objective interpretation of dobutamine effects on left ventricular wall motion. In addition, the time-velocity plot method provided automatic detection of peak velocity, timing, and duration of wall velocity changes over time. (J Am Soc Echocardiogr 1999;12:22-31.)

detect anomalies in ischemic myocardial segments during dobutamine infusion1 because it is independent of the translational motion of the heart.2,3 This is a major issue because dobutamine stress echocardiography is limited by subjective visual analysis of segmental wall thickness, and dobutamine enhances cardiac motion in the chest. MVG has been previously assessed by 1 method3-6 measuring velocities across the thickness of the myocardium (thickness-velocity plots) at the time of maximal color brightness assumed to correspond to peak endocardial velocity. MVG was calculated as the slope of the regression line between wall velocities and wall thickness.This process must be repeated in each segment during systole with the use of 2-

Journal of the American Society of Echocardiography Volume 12 Number 1

dimensional color DMI images or in each phase of the cardiac cycle with the use of M-mode images. Palka et al5 proposed to limit calculations to a few selected cardiac phases to make this method less time-consuming. Moreover, manually selected velocity measurements may not be feasible when myocardial walls are only partially color-encoded as in ischemic segments. In a recent report,6 although time-velocity plots were obtained from endocardial velocities, MVG was calculated from thicknessvelocity plots. We suggested that MVG could be automatically assessed throughout the cardiac cycle from velocity measurements parallel to the endocardial boundary to the left ventricular (LV) cavity on the one hand and parallel to the epicardial boundary on the other (time-velocity plots) with the use of Mmode recordings for optimal temporal resolution. With the use of this new assessment method, peak velocities were automatically determined in each phase of the cardiac cycle, and MVG was calculated as the difference between peak endocardial velocity and peak epicardial velocity divided by wall thickness. The purpose of this study was to compare MVG obtained from the time-velocity plot method with that obtained from the thickness-velocity plot method in quantifying dobutamine-induced changes in myocardial wall motion on normal hearts with the use of color M-mode DMI recordings. This aim entailed the following requirements: (1) definition of criteria for MVG assessment, (2) normal and stable hemodynamic parameters, and (3) choice of animals chronically instrumented for ethical reasons in the design of a methodological study.

METHODS Animal Preparation Eight chronically instrumented, conscious adult mongrel dogs weighing 22 to 28 kg were investigated.As previously reported,7 dogs were instrumented for the measurement of arterial pressure, LV pressure, and LV dP/dt. All experiments were conducted when the dogs were healthy, apyretic, and had been trained to lie quietly on the experiment table. Examinations were performed at steady state before and during dobutamine infusion (10 µg/kg per minute for 15 minutes). Data were continuously recorded on a multichannel recorder (ES 2000, Gould Instruments, Inc).The ensuing experiments were approved by the institutional animal care and use committee and were performed in accordance with the official regulations of the French Ministry of Agriculture.

Pellerin et al 23

Echo-Doppler Equipment One- and 2-dimensional gray-scale and DMI images in the velocity mode were recorded with an Acuson 128 XP 10 ultrasound scanner (Acuson Inc, Mountain View, Calif) with a 4.0-MHz probe. A regional enhancement selection box was always used to optimize the frame rate for depth and the amount of bursts per scan line and to obtain better color encoding.With the box size used in this study, the frame rate was up to 30 to 40 images per second. Twodimensional and M-mode standard echo and color DMI recordings were stored on half-inch Pal S-VHS videotape. Echocardiographic Examination Dogs were lying in the left lateral recumbent position, and a standard gray-scale transthoracic echocardiographic 2dimensional short-axis view was imaged at the level of papillary muscles.8 The M line was located perpendicular to the septal and posterior walls. Wall thickness and LV fractional shortening (%) were calculated.9 Images were recorded simultaneously with an electrocardiographic lead and a phonocardiogram (Cambridge Heart, Inc, Cambridge, Mass). DMI Examination Because of the thoracic V-shaped morphology and superficial heart of the dogs, optimal definition of myocardial posterior wall boundaries was easily obtained on the grayscale image.The regional enhancement selection box was located on the posterior wall beyond the mitral valve. We chose the velocity scale allowing recording of the lowest velocities without aliasing. Color M-mode DMI images were recorded continuously and selected during expiration. On the freeze-frame M-mode image, DMI and grayscale displays were successively and separately obtained. The gray-scale image was used to precisely locate LV diastolic and systolic endocardial and epicardial boundaries with calipers.These calipers could also be seen on the colored image to define endocardial and epicardial boundaries.10 Color M-mode recordings of the LV posterior wall were performed on a 100 mm/s strip chart. Computerization of M-Mode Color Doppler Wall Encoded Images The freeze-frame images were digitized on-line from the video output through an acquisition card provided with the machine (QV 100 Aegis, Acuson, Inc, Mountain View, Calif).The digitized images were transferred into a personal computer (Power Macintosh 9500/200, Apple Computer Inc, Cupertino, Calif), to be processed by the NIH image software (National Institutes of Health, Bethesda, Md) modified for automatic acquisition of wall velocities. The format of the digitized image was 768 longitudinal pixels × 576 vertical pixels. The pixel size was typically approximately 3.8 ms × 0.06 mm, and measured velocities repre-

24 Pellerin et al

Journal of the American Society of Echocardiography January 1999

Figure 1 Two myocardial velocity gradient (MVG) assessment methods on color M-mode Doppler myocardial image recording of the left ventricular posterior wall (LVPW). Upper panel, Color M-mode Doppler myocardial image recording of the LVPW in the short-axis view at baseline. Left lower panel, Thickness-velocity plot method. Mean myocardial velocities were measured every 0.06 mm along a vertical line traced across the thickness of the myocardium at the time of maximal color brightness during rapid ventricular filling. MVG was assessed by the slope of the regression line between mean velocities and wall thickness. Right lower panel, Time-velocity plot method. Mean myocardial velocities were measured every 3.8 ms throughout the cardiac cycle along a segmented line traced in the subendocardium parallel to the endocardial boundary and along another segmented line traced in the subepicardium parallel to the epicardial boundary. Peak value of mean myocardial wall velocities and time intervals were automatically singled out in each phase of the cardiac cycle and in each layer. Early systole and rapid ventricular filling were easily singled out by the phonocardiogram. Endocardial and epicardial boundaries were precisely located by use of the gray-scale image. In each phase of the cardiac cycle, MVG was assessed by the difference between subendocardial and subepicardial peak velocities divided by wall thickness. Strip chart 100 mm/s. PCG, phonocardiogram; ECG, electrocardiogram.

Journal of the American Society of Echocardiography Volume 12 Number 1

Pellerin et al 25

Figure 2 Myocardial velocity gradient (seconds-1) during 8 phases of the cardiac cycle at baseline (open bars) and during dobutamine infusion (solid bars). Myocardial velocity gradient was assessed by the thickness-velocity plot method (top) and by the time-velocity plot method (bottom). Both methods showed a significant increase in myocardial velocity gradient induced by dobutamine during preejection, systole, and rapid ventricular filling. Data are presented as mean ± SD (n = 8). *P less than .05 between baseline and dobutamine infusion. A, Atrial contraction; D, diastole; E, rapid ventricular filling; IR, isovolumic relaxation; PEP, preejection period; S, systole.

sented the mean velocity of several myocardial fibers contained in pixels (lowest measurable velocity 2 mm/s). Each M-mode color image represented 1.2 Mb without compression and was stored on a magneto-optical driver (Sierra 1.3 Gb, Pinnacle Micro, Inc, Irvine, Calif). Image Processing According to the Doppler principle, velocities of wall motions toward the transducer are positive (displayed in red) and away from the transducer are negative (displayed in blue). Successive changes in direction of motion are seen as color reversals featuring color strips. The color strips of the LV posterior wall M-mode recordings have been validated from hemodynamics as a reliable guide to define cardiac phases11 from the onset of the Q wave on the electrocardiogram: preejection period; early, mid, and

end systole; isovolumic relaxation; rapid ventricular filling; mid diastole; and atrial contraction. Except for the preejection period, wall velocities had the same direction within each cardiac phase. We also used a phonocardiographic checking of the heart sounds. The DMI analysis system that we developed was used to automatically convert color-coded velocities of the myocardium into velocity estimates from the color bar data at the left side of the DMI image. Two procedures were performed to assess MVG. With the use of the thickness-velocity plot method, mean myocardial velocities were automatically measured along vertical lines perpendicular to the endocardial border across the thickness of the myocardium at the time of maximal color brightness in all phases of the cardiac cycle and at the time of inward motion during the preejection

26 Pellerin et al

Journal of the American Society of Echocardiography January 1999

Figure 3 Calculation of myocardial velocity gradient by use of the thickness-velocity plot method during rapid ventricular filling at baseline and during dobutamine infusion. The gray-scale M-mode image was used to precisely locate left ventricular diastolic and systolic endocardial and epicardial boundaries. These boundaries can be seen on the colored M-mode Doppler myocardial image; myocardial velocities were measured across the thickness of the myocardium every 0.06 mm. Myocardial velocity gradient was assessed by the slope of the regression line between wall velocities (cm/s) and wall thickness (mm). Dobutamine induced a significant increase in myocardial velocity gradient related to a greater rise in endocardial than in epicardial velocities.

period.12 MVG was expressed as a positive value when endocardium was moving faster than epicardium, as suggested by Palka et al.4 Velocity was automatically measured every 0.06 mm throughout wall thickness, showing progressive changes in myocardial velocities. A correction algorithm for small non–color-encoded areas detected when differences between adjacent velocities exceeded 10% of the maximal value of the velocity scale was developed to improve reproducibility. These black zone velocities were excluded for MVG assessment. With the use of the new time-velocity plot method, mean myocardial velocities were automatically measured every 3.8 ms throughout the cardiac cycle between the onset of 2 consecutive Q waves along a segmented line traced in the subendocardium defined as the myocardial layer located below the endocardial boundary, along another segmented line traced in the subepicardium parallel to the epicardial boundary, and along a third segmented line traced in the mid wall, halfway between the 2 previous lines (Figure 1). Peak value of mean myocardial wall velocities and time intervals were automatically singled out in each phase of the cardiac cycle and in each layer. All measurements were averaged over 3 cardiac cycles. DMI Parameters Myocardial wall velocities (centimeters per second), MVG (seconds-1), and time measurements (milliseconds) were

studied at baseline and at steady state during dobutamine infusion. From the thickness-velocity plot method, MVG was assessed by the slope of the regression line between mean velocities and wall thickness. From the time-velocity plot method, MVG was assessed by the difference between subendocardial and subepicardial peak velocities divided by wall thickness. For both procedures, MVG was considered present in each phase of the cardiac cycle if its mean value was significantly different from zero and if endocardial and epicardial velocities were significantly different. Time measurements consisted of strip duration and time to peak velocity from the Q wave of the electrocardiogram in each color strip. Other Parameters Heart rate (beats/min), LV shortening fraction (%), mean arterial pressure (mm Hg), first derivative of LV pressure (mm Hg/s), LV end-diastolic pressure (mm Hg), and stroke volume (mL) were measured. Statistical Analysis Data are expressed as mean ± SD. Bland and Altman analysis13 was used to assess agreement in MVG between the 2 assessment methods. Significance of differences in MVG, in subendocardial, mid wall, subepicardial peak velocities, and in hemodynamic parameters mean values between

Journal of the American Society of Echocardiography Volume 12 Number 1

Pellerin et al 27

Table 1 Myocardial velocity gradient (seconds-1) in 8 phases of the cardiac cycle at baseline and at steady state during dobutamine infusion assessed by the 2 methods Thickness-velocity plot method Baseline

PEP Early S Mid S End S IR E Mid D A

2.64 2.14 1.90 1.37 0.76 3.06 0.69 1.82

± ± ± ± ± ± ± ±

Dobutamine

0.83 0.59 1.06 0.57 1.22 1.12 1.07 1.05

4.05 6.08 5.31 2.44 0.95 7.82 0.84 2.55

± ± ± ± ± ± ± ±

Time-velocity plot method P value

0.81 2.20 2.95 0.53 1.30 2.58 0.92 1.66

Baseline

<.05 <.01 <.01 <.01 NS <.01 NS NS

3.24 2.61 2.92 1.93 0.72 3.35 0.64 2.60

± ± ± ± ± ± ± ±

Dobutamine

0.94 1.82 1.97 0.68 1.03 1.35 0.88 1.26

4.64 7.12 5.87 3.83 0.80 6.85 0.81 2.89

± ± ± ± ± ± ± ±

P value

1.07 1.80 2.34 1.60 1.11 2.39 0.92 1.66

<.05 .001 <.01 <.01 NS <.01 NS NS

P value, Significance of differences in myocardial velocity gradient mean values between baseline and dobutamine infusion (paired t test); A, atrial contraction; D, diastole; E, rapid ventricular filling; IR, isovolumic relaxation; PEP, preejection period; S, systole; NS, not significant.

Table 2 Subendocardial, mid, and subepicardial peak velocities (cm/s) throughout the cardiac cycle at baseline and at steady state during dobutamine infusion measured with the time-velocity plot method Baseline Subendocardium

PEP Early S Mid S End S IR E Mid D A

4.7 4.2 6.1 3.9 –3.1 –10.0 –1.5 –2.3

± ± ± ± ± ± ± ±

0.4 1.8 1.4 1.1 1.7 1.1 1.5 1.4

Mid wall

3.3 3.0 4.7 2.9 –2.7 –8.3 –1.7 –0.8

± ± ± ± ± ± ± ±

0.7 1.9 1.2 1.1 1.2 2.7 1.0 1.9

Dobutamine infusion Subepicardium P value

2.1 2.1 2.5 1.5 –2.2 –5.8 –0.9 –0.2

± ± ± ± ± ± ± ±

0.4 1.8 0.9 1.4 2.3 1.4 1.4 1.8

<.01 <.001 <.001 <.001 NS <.001 NS <.01

Subendocardium

8.4 9.4 10.6 7.2 –4.3 –14.6 –2.7 –3.4

± ± ± ± ± ± ± ±

1.2* 2.2* 2.5* 1.2* 1.6 3.7* 1.6 1.7

Mid wall

6.4 7.5 6.8 4.6 –3.8 –11.6 –2.0 –1.6

± ± ± ± ± ± ± ±

1.4* 4.2* 1.6* 1.3* 1.5 3.8* 1.3 1.4

Subepicardium

4.5 2.9 3.2 2.4 –3.3 –6.0 –1.9 –1.0

± ± ± ± ± ± ± ±

0.9* 2.0 0.7 1.0 1.9 2.6 1.4 0.8

P value

<.01 <.001 <.001 <.001 NS <.001 NS <.001

P value, Significance of differences in peak velocity mean values between the 3 layers in each phase of the cardiac cycle (analysis of variance); A, atrial contraction; D, diastole; E, rapid ventricular filling; IR, isovolumic relaxation; PEP, preejection period; S, systole; NS, not significant. *Significance of differences in subendocardial, mid wall, and subepicardial peak velocities mean values between baseline and dobutamine infusion (paired t test). P less than .05.

baseline and dobutamine infusion was assessed by a paired t test. MVG mean value during each phase of the cardiac cycle was compared with zero by use of the t test. Influence of heart rate changes on MVG changes induced by inotropic modulation was assessed by a 3-way analysis of variance (SAS Software). In each phase of the cardiac cycle at baseline and during dobutamine infusion, significance of differences in peak velocity and in time to peak velocity mean values between the 3 layers were assessed by analysis of variance and Bonferroni test, successively. A value of P less than .05 was considered significant. Interobserver and Intraobserver Variabilities From random selection, 1300 mean myocardial velocities and 40 MVG were measured by 2 independent observers and within a 3-week time by 1 observer. They were expressed as a percent difference between the 2 values divided by the mean of both values. Interobserver variabilities in myocardial velocities and in MVG were 0.01% ± 0.21% and 0.09% ± 0.27%, respectively (not significant),

and intraobserver variabilities were 0.01% ± 0.19% and 0.08% ± 0.30%, respectively (not significant).

RESULTS The feasibility of color M-mode DMI was 100%.There was a significant positive correlation between MVG obtained from the thickness-velocity plot method and from the time-velocity plot method (r = 0.82, P < .0001).There was a close agreement between the 2 MVG assessment methods. The bias (mean difference in MVG between the 2 methods) of all measurements was –0.17 ± 9.71 seconds-1. MVG values obtained by the time-velocity plot method were significantly higher than those obtained by the thickness-velocity plot method during the preejection period, systole, and atrial contraction. According to our criteria and for both methods, MVG was present during the preejection period, sys-

Journal of the American Society of Echocardiography January 1999

28 Pellerin et al

Table 3 Color strip durations (ms) and time to peak velocity from Q wave (ms) in subendocardium and subepicardium throughout the cardiac cycle at baseline and at steady state during dobutamine infusion measured with the time-velocity plot method Baseline

Dobutamine infusion

Time to peak velocity Strip duration

PEP Early S Mid S End S IR E Mid D A

43 51 83 47 37 96 207 49

± ± ± ± ± ± ± ±

9 17 26 9 7 37 52 8

Subendocardium

28 77 120 179 250 303 435 584

± ± ± ± ± ± ± ±

19 21 20 15 16 18 135 60

Time to peak velocity

Subepicardium

20 64 127 180 244 300 411 574

± ± ± ± ± ± ± ±

18 24 22 20 26 27 74 73

Strip duration

40 36 70 36 29 83 134 35

± ± ± ± ± ± ± ±

8 9 15 6 7 26 27 9

Subendocardium

23 55 116 150 199 250 348 442

± ± ± ± ± ± ± ±

Subepicardium

15 16 14 17 11 21 63 35

19 58 125 149 200 236 328 432

± ± ± ± ± ± ± ±

15 13 16 12 29 36* 64 35*

A, Atrial contraction; D, diastole; E, rapid ventricular filling; IR, isovolumic relaxation; PEP, preejection period; S, systole. *Significance of differences in time to peak velocity mean values between endocardial and epicardial layers in each phase of the cardiac cycle.

tole, rapid ventricular filling, and atrial contraction. Its values ranged from 1.37 to 7.82 seconds-1.When MVG was absent, its values ranged from 0.64 to 0.95 seconds -1 (Table 1). Dobutamine induced a significant increase in MVG during the preejection period, systole, and rapid ventricular filling (Figure 2) related to a greater rise in endocardial than in epicardial velocities. MVG Obtained from the Thickness-Velocity Plot Method The MVG values are shown in Table 1. At baseline, the regression line between wall velocities and wall thickness showed a harmonious decrease in velocities from endocardium to epicardium. Correlation coefficients ranged from 0.61 to 0.87 (P < .01). During dobutamine infusion, MVG increase was evidenced by steeper regression slopes (Figure 3). Correlation coefficients ranged from 0.66 to 0.94 (P < .01). MVG Obtained from the Time-Velocity Plot Method The MVG values and peak velocities in the 3 myocardial layers are shown in Tables 1 and 2, respectively. Both at baseline and during dobutamine infusion, the highest peak velocities in absolute values were found in the subendocardium during rapid ventricular filling. Dobutamine significantly increased subendocardial and mid wall peak velocities during systole and rapid ventricular filling and increased peak velocities in the 3 layers during the preejection period (Figure 4).The increase in subendocardial velocities caused by the dobutamine effect featured a plateau between early and mid systole.All of the above dobutamine-induced changes remained significant when

the influence of changes in heart rate was taken into account. Other Information Obtained from the TimeVelocity Plot Method In addition to MVG and peak velocities, this method provided time measurements shown in Table 3. Durations of LV ejection measured from color strips were 181 ± 17 ms at baseline and 161 ± 24 ms during dobutamine infusion (11% decrease). Diastole lasted 389 ± 75 ms at baseline and 280 ± 36 ms during dobutamine infusion (28% decrease). Time to peak velocities occurred significantly earlier in subepicardium than in subendocardium in rapid ventricular filling and atrial contraction during dobutamine infusion. General Findings Dobutamine induced a significant increase in heart rate (from 99 ± 12 to 130 ± 9 beats/min), LV shortening fraction (from 44% ± 6% to 69% ± 8%), first derivative of LV pressure (from 3805 ± 307 to 5220 ± 310 mm Hg/s), and stroke volume (from 19 ± 2 to 24 ± 2 mL). No significant change was noted in mean arterial pressure (from 115 ± 4 to 115 ± 3 mm Hg) and LV end-diastolic pressure (from 6.5 ± 0.7 to 6.4 ± 0.6 mm Hg).

DISCUSSION DMI has been shown to be reliable for addressing ventricular wall mechanical events related to cardiac function through measurements of wall velocities.4,14-18 MVG was first described by McVeigh and Zerhouni19 with the use of magnetic resonance tag-

Journal of the American Society of Echocardiography Volume 12 Number 1

Pellerin et al 29

Figure 4 Subendocardial and subepicardial peak velocities (cm/s) measured throughout the cardiac cycle at baseline (open bars) and during dobutamine infusion (solid bars) by use of the time-velocity plot method. Subendocardial peak velocities (top) were significantly increased by dobutamine during the preejection period, systole, and rapid ventricular filling. In contrast, subepicardial peak velocities (bottom) were only significantly increased by dobutamine at the time of the inward motion during the preejection period. Data are presented as mean ± SD (n = 8). *P less than .05 between baseline and dobutamine infusion. A, Atrial contraction; D, diastole; E, rapid ventricular filling; IR, isovolumic relaxation; PEP, preejection period; S, systole.

ging and by Fleming et al20 with the use of color Mmode DMI. MVG has been shown to represent regional wall motion independent of overall cardiac translation in the chest.2,3 In the current study, MVG was calculated on conscious dogs in the absence of any confounding effect as assessed from hemodynamics, providing reference values for normal hearts. A visual qualitative analysis of color Doppler myocardial images cannot show small velocity changes in myocardial wall thickness and an automatic quantification of velocities is mandatory to calculate MVG.Wall motion during myocardial contrac-

tion reflects the rate of increase in wall thickness. Showing faster endocardial than epicardial motion during myocardial contraction, our findings are consistent with literature.21 MVG can only be assessed with color DMI. Although myocardial velocities are easy to measure with calipers by use of spectral pulsed DMI, subendocardial and subepicardial velocities cannot be separately analyzed because the fixed sample volume overrides moving structures of several layers during the cardiac cycle. In previous studies with M-mode DMI record-

30 Pellerin et al

ings,4-6 MVG was assessed by the slope of the regression line between wall velocities and wall thickness. To our knowledge, this is the first time that criteria for its presence and that correlation coefficients are provided. These 2 MVG assessment methods each has its own advantages. The thickness-velocity plot method is time-consuming when applied to 8 phases of the cardiac cycle.Therefore, the analysis of 3 selected phases has been proposed.5 However, it relied on the assumption that visually selected maximal color brightness corresponds to peak endocardial velocity in all phases of the cardiac cycle, a questionable statement when myocardial color encoding is suboptimal. In addition to myocardial velocities and MVG assessed by both methods, the time-velocity plot method provided automatic detection of peak velocity and time to peak velocity in each layer during each cardiac phase.This method with automatic detection of peak velocity might turn out to be a precious advantage when suboptimal myocardial color encoding makes its visual determination questionable. Accordingly, MVG values obtained from this method were higher than those obtained from the thickness-velocity plot method. Durations and sequential changes in wall motion over time are also precisely specified by using the time-velocity plot method, making the analysis more refined and interfacing the new velocity data with previous physiologic knowledge on duration events. The duration of ejection measured from color strips is consistent with that previously measured in dogs.22 Dobutamine-induced changes in MVG have been studied with the thickness-velocity plot method in systole.1 Interestingly, the time-velocity plot method showed that peak velocities simultaneously occur in endocardium and in epicardium during dobutamine infusion in systole but not in diastole. Thus information from both methods was complementary. MVG has been recently shown to detect ischemic myocardium during dobutamine infusion in the anteroseptal and posterior segments, whereas endocardial velocities and visual interpretation failed to clearly demonstrate the differing responses between nonischemic and ischemic segments.1 Although dobutamine increases both myocardial wall motion and the translational motion of the heart, MVG has the specific advantage to reflect wall motion independent of overall cardiac motion. The reason for MVG increase after dobutamine is not clearly elucidated.The positive inotropic effect of dobutamine is largely caused by β-1 adrenoreceptor stimulation supplemented by activation of cardiac β-2 and α-1 adrenoreceptors. Coronary blood flow increases

Journal of the American Society of Echocardiography January 1999

consistently with dobutamine, regardless of whether or not coronary artery obstruction is present.23,24 On the one hand,in normal nonischemic myocardium, endocardial and epicardial blood flows have been shown to increase proportionally with unchanged endocardial/epicardial flow ratio.25 On the other hand, an additional loss of high-energy phosphates in the subepicardium induced by dobutamine only occurred in ischemic myocardium.26,27 Further studies are required to elucidate the systolic plateau and the temporal dissociation in time to peak velocity occurring during rapid ventricular filling between the 3 layers. Study Limitations The rotational component of cardiac translation is likely to have minimal effects on the mid part of the left ventricle.28 The linear regression line between wall velocities and wall thickness cannot be explained related to the heterogeneous nature of myocardial fiber orientation29 and sequence of contractile activation.30,31 The effects of loading conditions on wall velocities remain to be defined over a range of clinical situations. Application of our results to human beings will require studies in volunteers. Although color M-mode DMI recordings have a high temporal resolution, this modality cannot display all LV segments at this time because of the Doppler angle dependency. 3 Given the aim of the study, implanted length crystals were not mandatory. Implications MVG assessment has potential applications under conditions of asynchronous contraction such as may occur with transmural or subendocardial ischemia. On the basis of our results, Doppler myocardial imaging has the potential to enhance objective interpretation of stress echocardiograms. Cardiac wall tracking should facilitate such applications.32

CONCLUSIONS Our study showed that 2 complementary methods may be applied to MVG assessment and enabled objective interpretation of dobutamine effects on LV wall motion. For the first time, quantitative criteria have been used for assessing MVG, and its range values have been given in normal hearts at baseline and during dobutamine infusion. In addition to the thickness-velocity plot method, the time-velocity plot method only provides peak velocity during each cardiac phase, sequential changes in LV wall motion

Journal of the American Society of Echocardiography Volume 12 Number 1

over time, and time measurements as a frame for a new methodological approach.

REFERENCES 1. Tsutsui H, Uematsu M, Shimizu H, et al. Comparative usefulness of myocardial velocity gradient in detecting ischemic myocardium by a dobutamine challenge. J Am Coll Cardiol 1998;31:89-93. 2. Uematsu M, Nakatani S, Yamagishi M, et al. Usefulness of myocardial velocity gradient derived from 2-dimensional tissue Doppler imaging as an indicator of regional myocardial contraction independent of translational motion assessed in atrial septal defect. Am J Cardiol 1997;79:237-41. 3. Uematsu M, Miyatake K, Tanaka N, et al. Myocardial velocity gradient as a new indicator of regional LV contraction: detection by a two-dimensional tissue Doppler imaging technique. J Am Coll Cardiol 1995;26:217-23. 4. Palka P, Lange A, Fleming AD, et al. Age-related transmural peak mean velocities and peak velocity gradients by Doppler myocardial imaging in normal subjects. Eur Heart J 1996;17: 940-50. 5. Palka P, Lange A, Fleming AD, et al. Differences in myocardial velocity gradient measured throughout the cardiac cycle in patients with hypertrophic cardiomyopathy, athletes and patients with LV hypertrophy due to hypertension. J Am Coll Cardiol 1997;30:760-8. 6. Gorcsan J III, Deswal A, Mankad S, et al. Quantification of the myocardial response to low-dose dobutamine using tissue Doppler echocardiography measures of velocity and velocity gradient. Am J Cardiol 1998;81:615-23. 7. Berdeaux A, Ghaleh B, Dubois-Randé JL, et al. Role of vascular endothelium in exercise-induced dilation of large epicardial coronary arteries in conscious dogs. Circulation 1994; 89:2799-808. 8. Thomas WP, Gaber CE, Jacobs GJ, et al. Recommendations for standards in transthoracic 2-dimensional echocardiography in the dog and cat. J Vet Intern Med 1993;7:247-52. 9. Feigenbaum H, Popp RL, Wolfe SB, et al. Ultrasound measurements of the left ventricle: a correlative study with angiography. Arch Intern Med 1972;129:461-7. 10. Veyrat C, Pellerin D. Doppler myocardial imaging: methodological recommendations and role of computerization to optimize efficient applications. J Noninvas Cardiol 1997;1: 26-34. 11. Zamorano J, Wallbridge DR, Ge J, et al. Noninvasive assessment of cardiac physiology by tissue Doppler echocardiography: a comparison with invasive hemodynamics. Eur Heart J 1997;18:330-9. 12. Pellerin D, Cohen L, Larrazet F, et al. Preejectional LV wall motion in normal subjects using Doppler tissue imaging and correlation with ejection fraction. Am J Cardiol 1997;80: 601-7. 13. Bland JM, Altman DG. Statistical method for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307-10. 14. McDicken WN, Sutherland G, Moran CM, et al. Colour Doppler velocity imaging of the myocardium. Ultrasound Med Biol 1992;18:651-4.

Pellerin et al 31

15. Sutherland GR, Stewart MJ, Groundstroem KWE, et al. Color Doppler myocardial imaging: a new technique for the assessment of myocardial function. J Am Soc Echocardiogr 1994;7:441-58. 16. Miyatake K, Yamagishi M, Tanaka N, et al. New method for evaluating LV wall motion by color-coded tissue Doppler imaging: in vitro and in vivo studies. J Am Coll Cardiol 1995; 25:717-24. 17. Gorcsan J III, Gulati VK, Mandarino WA, et al. Color-coded measures of myocardial velocity throughout the cardiac cycle by tissue Doppler imaging to quantify regional LV function. Am Heart J 1996;131:1203-13. 18. Garcia MJ, Rodriguez L, Ares M, et al. Myocardial wall velocity assessment by pulsed Doppler tissue imaging: characteristic findings in normal subjects. Am Heart J 1996;132:64856. 19. McVeigh ER, Zerhouni EA. Noninvasive measurement of transmural gradients in myocardial strain with MR imaging. Radiology 1991;180:677-83. 20. Fleming AD, Wia X, McDicken WN, et al. Myocardial velocity gradients detected by Doppler imaging. Br J Radiol 1994;67:679-88. 21. Heyndrickx GR, Baig H, Nellens P, et al. Depression of regional blood flow and wall thickening after brief coronary occlusions. Am J Physiol 1978;234:H653-9. 22. Wallace AG, Mitchell JH, Skinner NS, et al. Duration of the phases of LV systole. Circ Res 1963;12:611-9. 23. Sakanashi M, Tomomatsu E, Takeo S, et al. Effect of dobutamine on coronary circulation and cardiac metabolism of the dog. Drug Res 1978;28:798-801. 24. Bendersky R, Chatterjee K, Parmley WW, et al. Dobutamine in chronic ischemic heart failure: alterations in LV function and coronary hemodynamics. Am J Cardiol 1981;48:554-8. 25. Liang CS, Yi JM, Sherman LG, et al. Dobutamine infusion in conscious dogs with and without acute myocardial infarction: effects on systemic hemodynamics, myocardial blood flow and infarction size. Circ Res 1981;49:170-80. 26. Guth BD, Schulz R, Heusch G. Time course and mechanisms of contractile dysfunction during acute myocardial ischemia. Circulation 1993;87:IV-35-42. 27. Zhang J, Path G, Chepuri V, et al. Effects of dobutamine on myocardial blood flow, contractile function and bioenergetic responses distal to coronary stenosis: implications with regard to dobutamine stress testing. Am Heart J 1995;129:330-42. 28. Qi P, Thomsen C, Stahlberg F, et al. Normal LV wall motion measured with 2-dimensional myocardial tagging. Acta Radiol 1993;34:450-6. 29. Greenbaum RA, Yen Ho S, Gibson DG, et al. Left ventricular fibre architecture in man. Br Heart J 1981;45:248-56. 30. Gibson DG. Regional LV wall motion. In: Roelandt JRTC, Sutherland GR, Iliceto S, Linker DT, editors. Cardiac ultrasound. Edinburgh, UK: Churchill Livingstone; 1993:24151. 31. Pandian NG, Skorton DJ, Collins SM, et al. Heterogeneity of LV segmental wall thickening and excursion in 2-dimensional echocardiograms of normal human subjects. Am J Cardiol 1983;51:1667-3. 32. Cohen L, Pajany F, Pellerin D, et al. Cardiac wall tracking using Doppler tissue imaging. Proceedings of the International Conference on Image Processing. 1996;3:295-8.