Abstract: We present a new approach to analyze the deformation of the left ventricle of the heart, based on a parametric model that gives a compact representation of a set of points in a 3D image. We present four different approaches to tracking surfaces in a sequence of 3D cardiac images. Following tracking, we then infer quantitative parameters which are useful for the physician, such as the variation of volume and wall thickness during a cardiac cycle, the ejection fraction or the twist component in the deformation of the ventricle. We explicit the computation of these parameters using our model. Experimental results are shown in time sequences of two kinds of medical images, Nuclear Medicine and X-Ray Computed Tomography (CT).