Extraction Axis Of Three-Dimensional Blood Vessel Images Based On Energy Constraint Equation

With the continuous improvement of medical diagnostic tools and the popularity of applications such as magnetic resonance imaging (MRI) and computer tomography (CT), medical images produced by them have been routinely used to assess the shape, function and state of the patients'blood vessel. The diameter, length and volume of blood vessel can help doctors diagnose, plan the implementation of surgery and treat patients, these parameters are important, but they are not easy to accurately calculate. Extraction axis of blood vessel is the first step in quantifying blood vessel and is also the most important step, it has been a great concern in the medical image processing.The methods of extracting axis of three-dimensional (3D) blood vessel images are usually thinning method based on morphological, distance transform and model method etc. This paper focuses on analysis and research of the important roles of thinning algorithm and distance transform during extracting axis of blood vessel. By structuring the energy constraint equation based on blood vessel characteristics and combining with active contour modeling (snake modeling), the method of extracting axis of 3D blood vessel images based on the energy constraint equation was proposed. We can get the initial skeleton curves of blood vessel images by two ways, one is to using thinning algorithm, and the other way is that some initial points will be manually pre-selected, the successive line segments between the two points are projected to the vessel surface, the initial skeleton curve will be obtained on vessel surface finally. The former approach has no manual participation and the last one is more time-saved. With the role of energy constraint equations, the initial skeleton obtained by either way of the two can continuously iterate and evolve, eventually converge to the exact location of blood vessel axis.Due to minimizing engery constraint equation can be regarded as function optimization, thereof, we have introduced intelligence optimization algorithm during extracting axis of blood vessel. In this paper, we emphasisly reseached and improved partice swarm optimization algorithm, and proposed the concept"centroid"of partice swarm. Finally, the developed PSO was applied to extraction axis of blood vessel, and the final axis is more accurate.