Semi-automatic border detection of cardiac cavity in echocardiography video

In this paper, a semi-automatic border detection of cardiac cavity in two-dimensional short axis echocardiography video is proposed. The method used b-spline and optical flow to detect, reconstruct the border and tracking the border for every frame in echocardiography video. The first step is applying median filtering to eliminate noise. The second step is determining the initial boundary of the cardiac cavity in the first frame. The third step is using optical flow method to detect cardiac cavity and reconstruct the imprecise border. This technique can detect border of the cardiac cavity in the echocardiography video very precisely.