I. Introduction
As an indispensable diagnosis-aid technology, ultrasound (US) examination is of increasing importance in recent years [1]. It provides the B-scan with low cost and no radiation in real time [2]. The use of US would even increase in daily medicine and healthcare routines in the future. Traditional two-dimensional (2-D) US can dynamically display 2-D images of the region of interest [3]. However, the lack of 3-D anatomy information makes the diagnosis result heavily dependent on the subjective experience of the doctor. The 3-D US can overcome this serious limitation [4]–[9]. 3-D US imaging can be realized with three main steps: first, collect the raw B-scans labeled with corresponding spatial information; second, reconstruct the 3-D volume; and finally, render and display the 3-D volume. Every step has a significant influence on the accuracy of 3-D imaging. Except for systems using 2-D arrays, 3-D US images are reconstructed from raw B-scans, with their respective spatial information. Therefore, the scanning manners (especially the skills for manipulating the probe) and precise recording of position and orientation of B-scans are very important to ensure correct 3-D reconstruction. Only based on accurate 3-D US imaging data, 3-D image analysis [10], [11] can be further realized.