Comparison of fusion techniques for 3D+T echocardiography acquisitions from different acoustic windows

Abstract

We propose a new method to combine real-time 3D echocardiography acquisitions from different apical windows. It consists in registering the volumes using only the image data, without external positioning sensors, and combining their intensity values, without requiring a feature extraction method. Registration shows reasonable quantitative results on phantom data and visually acceptable results on in vivo data. We present two new methods for the fusion process: generalized averaging and multiview deconvolution. The latter, which performed best, improved the myocardium contrast by +49.67% and the signal-to-noise ratio by +4.61 dB. These improvements lead to more accurate quantification of cardiac function