Automated Assessment of Noninvasive Filling Pressure Using Color Doppler M-Mode Echocardography zyxw NL Greenberg, MS Firstenberg, LA Cardon, J Zuckerman, BD Levine, MJ Garcia, JD Thomas The Cleveland Clinic Foundation, Cleveland, USA Abstract zyxwvutsr Assessment zyxwvutsrqpo of left ventricular filling pressure usually requires invasive hemodynamic monitoring to follow the progression of disease or the response to therapy. Previous investigations have shown accurate estimation of wedge pressure using noninvasive Doppler information obtained from the ratio of the wave propagation slope from color Mmode (CMM) images and the peak early diastolic filling velocity from transmitral Doppler images. zyxwvut This study reports an automated algorithm that derives an estimate of wedge pressure based on the spatiotemporal velocity distribution available from digital CMM Doppler images of L Vfilling. 1. Introduction Figure 1: Color M-mode Doppler image demonstrating Assessment of left ventricular filling pressure is a common clinical problem in patients with established flow (generally color-coded) propagating from the left heart disease and usually requires invasive hemodynamic atrium (LA) the tMV) into the left ventricle (LV). monitoring to follow the progression of disease or the response to therapy. Several non-invasive Doppler echocardiographic indices have been proposed to estimate left ventricular (LV) filling pressures measured either directly or from pulmonary capillary wedge pressures during right heart catheterization.[ 1,2] A recently validated index using the ratio of early transmitral filling velocities (E) to the early diastolic LV flow propagation velocity (Vp) as measured using color M-mode Doppler echo. zyxwvutsrqpon [ 31 zyxwvutsrqpo A recent novel application of color Doppler echocardiography has been assessment of left ventricular filling by color Doppler M-mode.[4] By placing a scanline from the mid-left atrium through the mitral valve to the mid-left ventricle, a spatiotemporal velocity map of diastolic filling can be obtained. In a color Doppler M mode image, the x-dimension is time (t) increasing to the right and the y dimension is depth (s) from the transducer increasing down the image. In epicardial or transthoracic Doppler images as shown in Figure 1, blood flow across the mitral valve is towards the transducer, from the left atrium (LA) into the left ventricle (LV). 0276-6547/01 $17.00 0 2001 IEEE 601 One common parameter extracted from these images is the flow propagation velocity (v,,), given by the slope of the leading edge of the color Doppler early filling (E- wave) as it passes from left atrium to left ventricle as shown in Figure 1. Note that this propagation velocity is distinct from the velocities within the propagating wave (the velocities recorded by a pulsed Doppler sample volume between the mitral leaflets), which we term the "component" velocities. Brun et al. found in patients with dilated cardiomyopathy, propagation velocity was only 20% of the peak &wave (component) velocity, but these were nearly equal in normal controls.[5] They hrther demonstrated an inverse relationship between flow propagation velocity and ventricular relaxation time constant, ?. Stugaard et al. has proposed an alternative parameter of flow propagation, the temporal delay between the passage of flow at the mitral leaflet level to the ventricular apex.[6] This index, which can be obtained in a semi-automated manner from digitally output color M-mode images, was shown to be a sensitive measure of changes in ventricular filling induced by balloon angioplasty o f t he left anterior descending coronary artery. Computers in Cardiology 2001;28:601-604.