PSO Algorithm: A Case Study

The PSO algorithm was introduced in 1995 by Kennedy and Eberhart. In earlier days PSO can be used to implement the simple function and extended to real time applications such as medical image processing, industrial applications, satellite image processing etc. PSO is based on swarm intelligence concept. Swarm Intelligence is an artificial intelligence based on the collective behaviour of decentralized, self-organized systems in order to obtain the optimized results. Swarm is composed of collection of individuals known as particles. This particle poses the homogeneous characteristics. The local interaction of the particles is based on simple rules. Each particle can be self-organised. In this paper the process is divided into two stages: Pre processing and segmentation uses PSO.

Preprocessing is a method to remove the noise, film artifacts in MRI image. The MRI image consists of some film artifacts [1] such as patient

Step2: Fitness value of each particle in the population can be computed based on the following objective function in equation (6).

F(i,j)=1 / (1+X(i,j)2) (6)
Where,
F(i,j) denotes the fitness function
X(i,j) is the preprocessed image.

Step3: Initialize the velocity and position of the particles. Calculate the pBest value of each particle and consider as the initial position. Initially velocity is declared as zero and the position will be remains same as the intensity value of each pixel.

Step4: calculate the gBest which can be calculated by means of the pBest of each particle. The maximum pBest value among the particle can be considered as the gBest.

Step5: update the velocity of each particle using equation (7)

v [cp] = v[cp] + c1 * rand() * (pbest[p] - present[p]) + c2 * rand() * (gbest[p] - present[]) (7)

Where, v [cp] = current particle velocity, pbest[cp] =best fitness