START Conference Manager    

Left Ventricle Wall Motion Estimation in Echocardiographic Images

Nadia Souag, Saliha Tamrent, Med Mokhtar Djebbour and Abdelaziz Dahmani

The 2010 Summer Computer Simulation Conference (SCSC 10)
Ottawa, Canada, July 11-14, 2010


The aim of this paper is to estimate motion of the left ventricle in echocardiographic image sequences of normal heart. Cardiac wall kinetic analysis requires two steps: (i) detection of the left ventricle boundaries in the image sequence, (ii) myocardial wall motion estimation. We use the Level Set method to segment the left ventricle boundaries in 2D ultrasound images. This method implicitly represents the evolving contour by embedding it as the zero level of a hyper surface. The algorithm based on optical flow has been applied to estimate left ventricle wall motion. The optical flow at time t and location P(x,y,t) is defined as the velocity of the image point. The wall motion cardiac analysis method that we have proposed is based on the analysis of the evolution of the temporal trajectory of each pixel throughout the time t. To track the trajectory of each pixel, we used his velocity calculated using optical flow method. Assessment of left ventricular contraction is determined by variation time curves observed in each pixel, during a cardiac cycle. Each pixel trajectory was plotted using Hermite curves. Order to carry out the wall motion analysis, the left ventricle was divided up in a simple way into seven segments. This division was generated from three points positioned by the clinician.

START Conference Manager (V2.56.8 - Rev. 1182)