Swarm Intelligence Approach to 3D Medical Image Segmentation

In this paper we present a new idea for 3D medical image segmentation based on swarm intelligence and ant colony optimization. The methodology combines selected mechanisms running both mentioned artificial intelligence techniques, e.g. fitness-controlled motion of virtual agents or stigmergy. Foundations of the algorithm are described along with its implementation specification, simulations, results and their analysis also in terms of clarifying the parameterization. Several parameters are introduced and verified in terms of their influence on the method performance. The experiments rely on the segmentation of spleen in computed tomography studies. We also formulate some conclusions on possible ways for the algorithm future development.