Three dimensional reconstruction of the left ventricle based on the Delaunay triangulation

The aim of this paper is to develop software for three dimensional (3D) reconstruction of the left ventricle (LV) from two-dimensional (2D) echographic images. Using a rotating probe combined with an ECG gating technique we acquire transthoracic images database (20 images per ECG cycle for 9 rotations). Speckle noise is the major difficulty that arises in echocardiographic image processing. We have used an anisotropic diffusion to reduce the presence of the speckle. The LV boundaries are plotted manually using the PC mouse, and are sampled to reduce the amount of stocked data. Since the rotation step of probe is large (20/spl deg/), the spatial resolution is poor so additional boundaries are interpolated between acquired image contours. We have used the three-linear interpolation method. We can reconstruct 3D images of the left ventricle of apical long axis view of the same cardiac phase. The Delaunay triangulation method has been used to join pattern contours and interpolated points.