ABSTRACT
Echocardiography has become prevalent in the diagnosis of heart disease not only because its inherent characteristics of being economic, non-invasive, portable, and easy to operate, but also because it can depict the moving heart in real time, revealing dynamic information of health status at any cross-section of the heart in vivo. This chapter reviews the basic principles of this significant ultrasonic imaging tool of echocardiography, its application and the state-of-the-art approaches for classification of 3D echocardiograms. In particular, the attention will be focused on the work applying deep learning convolutional neural network (CNN) to the determination of viewpoint upon which the video images of interest are acquired from, aiming at assisting clinicians in the diagnosis of heart disease with high accuracy. In addition, hand-crafted approaches of 3D SIFT and 3D KAZE are elaborated.
