3D echocardiography data processing for wall motion estimation

A method for the estimation of motion of ventricular structures based on 3D echo data processing is presented. The Lucas-Kanade algorithm has been utilized and implemented in a multi-step version. The algorithm is based on the recognition of characteristic points (feature points) within the volume and to track them during motion. The extrapolation of the motion field to the entire volume has been performed by interpolating the estimated velocity vectors corresponding to the feature points by using a model for diffusion and regularization of the field. Results have been obtained on synthesized data and in vivo 3D echo data.