Analysis of inter-system variability of systolic and diastolic intraventricular pressure gradients derived from color Doppler M-mode echocardiography

Abstract

Assessment of intraventricular pressure gradients (IVPG) using color Doppler M-mode echocardiography has gained increasing interest in the evaluation of cardiac function. However, standardized analysis tools for IVPG quantification are missing. We aimed to evaluate the feasibility, the test-retest observer reproducibility, and the inter-system variability of a semi-automated IVPG quantification algorithm. The study included forty healthy volunteers (50% were men). All volunteers were examined using two ultrasound systems, the Philips Epiq 7 and the General Electric Vivid 6. Left ventricular diastolic (DIVPG) and systolic (SIVPG) intraventricular pressure gradients were measured from the spatiotemporal distribution of intraventricular propagation flow velocities using color Doppler M-mode in standard apical views. There was good feasibility for both systolic and diastolic IVPG measurements (82.5% and 85%, respectively). Intra and inter-observer test-retest variability measured with the intraclass correlation coefficient were 0.98 and 0.93 for DIVPG respectively, and 0.95 and 0.89 for SIVPG respectively. The inter-system concordance was weak to moderate with Lin's concordance correlation coefficient of 0.59 for DIVPG and 0.25 for SIVPG. In conclusion, it is feasible and reproducible to assess systolic and diastolic IVPG using color Doppler M-mode in healthy volunteers. However, the inter-system variability in IVPG analysis needs to be taken into account, especially when using displayed data.

Figure 1

Image acquisition and processing diagram.

Figure 2

Color Doppler M-mode recordings and processing of systolic and diastolic intraventricular pressure gradients (IVPG). The spatiotemporal velocity profile along the streamline from the left ventricular apex to the aortic annulus in systole and from the mitral annulus to apex in diastole, acquired by color Doppler M-mode, was used to compute systolic and diastolic IVPG after an automated dealiasing and smoothing process. Anatomic streamline markers were positioned at the mitral (red line) and aortic (green line) annulus. Intraventricular pressure difference curves were obtained during the entire cardiac cycle, using the Bernoulli equation. Peaks systolic and diastolic IVPG (black arrows) were averaged on three consecutive cycles.

Figure 3

Adjusting the grayscale gain at 60% provided the best compromise between color flow mapping and 2D gray-scale definition level of the left ventricular apex and base. The graphic (left panel) shows that a grayscale gain adjustment above 60% was responsible for a growing decrease in systolic (red line) and diastolic (blue line) intraventricular pressure gradients (IVPG). Data are expressed as a mean percentage decrease of peak IVPG from the initial value measured at 40% of the grayscale gain and averaged on fifteen different color Doppler M-mode acquisitions in systole and diastole. The right panel shows diastolic IVPG acquisitions made at a different grayscale gain (40%, 60%, 80%, and 90%).

Figure 4

Lin’s concordance correlation coefficient (CCC) and Bland Altman analysis for diastolic (DIVPG) and systolic (SIVPG) intraventricular pressure gradients measurements between system A and system B. The line of perfect concordance is the line of identity (y = x); the reduced major axis corresponds to the regression line between the two measurements.

Figure 5

Lin’s concordance correlation coefficient (CCC) between the pixel-to-pixel velocity value comparison extracted from the raw proprietary velocity data and the displayed velocity data estimated from the look-up table, on one sample of color-Doppler M-mode acquisition for both systems A and B. The line of perfect concordance is the line of identity (y = x); the reduced major axis corresponds to the regression line between the two measurements.

Figure 6

Effects of temporal displayed resolution changes on systolic (SIVPG) and diastolic (DIVPG) intraventricular pressure gradients measurements. Peaks SIVPG and DIVPG were assessed at different temporal displayed resolution settings (95 Hz, 190 Hz, 285 Hz, 380 Hz, and 570 Hz) in ten randomly selected subjects scanned with system A. Data are expressed as a mean percentage decrease of peaks SIVPG and DIVPG from the initial value measured at 570 Hz. Arrows indicate the default temporal displayed resolutions of system A and system B.