A very active research is conducted on notion analysis. Most of the concepts, methods and assumptions are well established and lead to additional improvements in computer vision applications. Even in medicine where we have to deal with noisy data, low contrast structures and deformable objects, they bring new cues at all the processing stages. This paper emphasizes the specificities of this area and also the potential difficulties to face. A compilation of results is given aimed at the quantification of heart kinetics in Digital Subtraction Angiography (DSA). They illustrate the benefits of cooperative schemes such as motion based segmentation, moving object identification, three dimensional reconstruction and interpretation.