We present an original Radial Basis Functions-based multiphase level set approach for the segmentation of cardiac structures in echocardiography. The method relies on two main contributions. We first describe a distribution allowing for the modeling of the radiofrequency signal for both blood and myocardial regions. We then formulate the problem of segmenting several cardiac regions in echocardiography using a Maximum Likelihood framework based on the proposed distribution. We minimize the resulting functional using a RBF-based multiphase level set model. Results obtained on both simulation and data acquired in vivo demonstrate the ability of our method to segment myocardial regions in echocardiography imaging.