LABELING OF MR BRAIN IMAGES USING BOOLEAN NEURAL-NETWORK

This paper presents a knowledge-based approach for labeling two-dimens ional (2-D) magnetic resonance (MR) brain images using the Boolean neu ral network (BNN), which has binary inputs and outputs, integer weight s, fast learning and classification, and guaranteed convergence. The a pproach consists of two components: a BNN clustering algorithm and a c onstraint satisfying Boolean neural network (CSBNN) labeling procedure , The BNN clustering algorithm is developed to initially segment an im age into a number of regions. Then the segmented regions are labeled w ith the CSBNN, which is a modified version of BNN [13]. The CSBNN uses a knowledge base that contains information on image-feature space and tissue models as constraints, The method is tested using sets of MR b rain images. The regions of the different brain tissues are satisfacto rily segmented and labeled. A comparison with the Hopfield neural netw ork and the traditional simulated annealing method for image labeling is provided. The comparison results show that the CSBNN approach offer s a fast, feasible, and reliable alternative to the existing technique s for medical image labeling.