An artificial ant colonies approach to medical image segmentation.

The success of image analysis depends heavily upon accurate image segmentation algorithms. This paper presents a novel segmentation algorithm based on artificial ant colonies (AC). Recent studies show that the self-organization of ants is similar to neurons in the human brain in many respects. Therefore, it has been used successfully for understanding biological systems. It is also widely used in many applications in robotics, computer graphics, etc. Considering the features of artificial ant colonies, we present an extended model for image segmentation. In our model, each ant can memorize a reference object, which will be refreshed when it finds a new target. A fuzzy connectedness measure is adopted to evaluate the similarity between target and the reference object. The behavior of an ant is affected by the neighbors and the cooperation between ants is performed by exchanging information through pheromone updating. Experimental results show that the new algorithm can preserve the detail of the object and is also insensitive to noise.