MRI myocardial perfusion analysis is an important element for the ischemic heart disorder assessment. The spatio-temporal analysis of the myocardial raising during the first crossing of a contrast bolus allows to identify the ischemic or hypoperfused areas. Such an analysis requires an accurate tracking of the myocardium on the whole sequence and a robust segmentation to identify pathological and healthy regions inside the myocardium. In this paper, we present a semiautomatic tracking tool and a segmentation algorithm based on statistical tests and a recent concentration theorem. We show experimental results for both the tracking step and the segmentation of hypoperfused areas to confirm the availability of this setting.