Objective The synchronous characteristics of the motion in the four cardiac cavities is important to accurately describe the mechanical activity of the heart, and especially important for the diagnosis and evaluation of post surgery of cardiac resynchronization therapy, yet currently there is no good quantitative measurement method. The whole heart ultrasound image sequence is rich with heart motion information and the aim of this paper is to extract the synchronous characteristics of the four-chamber motion of the heart from this kind of image sequence. Methods The basic idea of this paper is to segment the four cavities of each frame image in the whole heart ultrasound image sequence, and consequently calculate the plane areas of the four cavities, then record these plane areas over time, and finally obtain the synchronous characteristics of the four cavities through the feature analysis of the area curves. In this paper, the segmentation method is based on active shape model (ASM), which may overcome the low spatial resolution of ultrasound image, low SNR and vague cavity boundaries. To further optimize the method, this paper proposes an improved method based on local variance of images. Results Experimental results using simulated and real ultrasound images show that the method proposed in this paper is feasible. Conclusions Based on the improved ASM segmentation method, a good segmentation result and more accurate synchronous characteristics of the four cardiac cavities can be achieved.