Unsupervised Brain Magnetic Resonance Image Segmentation Using HMRF-FCM Framework

In this paper, image segmentation of brain magnetic resonance (MR) image is addressed in an unsupervised framework. We propose a novel method considering the hidden Markov random field model (HMRF) to model the image class labels, which takes into account the mutual influences of neighbouring sites formulated on the basis of fuzzy clustering principle. By introducing the effective means to incorporate the explicit assumptions of the HMRF model into fuzzy clustering procedure, an efficient fuzzy clustering-type treatment is yielded. This combines the benefits from the spatial coherency modelling capabilities of the HMRF model, and the enhanced flexibility obtained by the fuzzy clustering algorithm, i.e. fuzzy c-means algorithm (FCM). The proposed HMRF-FCM segmentation framework is validated with noisy synthesis as well as brain MR images. We experimentally demonstrate the superiority of the proposed approach over the existing HMRF-EM framework applied to brain MR image segmentation.