Tracking medical 3D data with a parametric deformable model

This paper describes a method to fit and track surfaces. We suppose that we have already extracted from a 3D image some data defined by a set of points. The fitting step makes use of a superquadric model and solves an inverse problem for free-form deformations. This is then used for tracking surfaces in a time sequence of 3D images. We present different approaches to track surfaces in a sequence of 3D images. From the tracking, we deduce an estimation of motion and visualize a velocity map. This method is applied to 2 sequences of different kinds of medical images to track the cardiac left ventricle.