3D image segment method with forecasting capability

A new 3D image segmentation algorithm was proposed. This algorithm integrat es the improved Active Contour model and a new forecast algorithm. The fore cast algorithm is based on the peculiarity of 3D images i.e. the deformatio n of the contour along the space or time axis is continuous in 3D images. T he forecast algorithm analyzes the trend of the deformation of the contour in the segmented images, then prognosticates the location and the shape of the contour in the next image. Experiments on medical anatomic images show that the new algorithm can observably improve the accuracy of the segmentat ion and reduce the rime needed.