Real-time synchronous hardware architecture for MRI images segmentation based on PSO

Abstract

Particle Swarm Optimization (PSO) is a metaheuristic algorithm based optimization technique for continuous search problem. It is among the most used algorithms in various areas of application. Its popularity has exceeded the deferred-time problems to the real-time problems that require the use of embedded architectures. Many real-time applications include mobile robots and medical image processing has been widely developed and improved using PSO by many researchers. We have succeeded the implementation of the real-time segmentation of MRI medical images based on PSO algorithm in previous work. In this paper, we try to extend the work by adding a control unit that controls each task of the various blocks of the architecture. Therefore, the new obtained synchronous architecture of MRI images segmentation based PSO allows to save execution time and thus narrow the search procedure of the optimal threshold. The performance of the proposed synchronous hardware architecture is evaluated and validated using a set of MRI medical images.