ENHANCING LEFT-VENTRICULAR SHORT-AXIS ECHOCARDIOGRAPHIC CINELOOPS WITH A PULSE COUPLED NEURAL NETWORK

Observing the motion of endocardial borders in echocardiographic cineloops is an important component for both the qualitative and quantitative analysis of regional left ventricular function. Consequently, using the computer to enhance endocardial border pixels is an important first step both for annotating and eventual computer analysis of echocardiograms. Previous work has demonstrated the viability of pulse coupled neural networks for enhancing echocardiograms acquired using an acoustical contrast agent. In this study we investigate using pulse coupled neural networks as a visualization tool, highlighting, frame-byframe, the endocardial border in left-ventricular short-axis cineloops.