Recently, the Phase Field Method has shown to be a powerful tool for variational image segmentation. In this paper, we present a novel multi-phase model for probability based image segmentation. By interpreting the phase fields as probabilities of pixels belonging to a certain phase, we obtain the model formulation by maximizing the mutual information between image features and the phase fields. For optimizing the model, we derive the Euler Lagrange equations and present their efficient implementation by using a narrow band scheme. We present experimental results on segmenting synthetic, medical and natural images.