Hybrid Particle Swarm Optimization and Its Application to Multimodal 3D Medical Image Registration

Chen-Lun Lin,Yen-Wei Chen,Aya Mimori

doi:10.1155/2012/561406

Chen-Lun Lin, Yen-Wei Chen + Show 1 more

Open Access

https://doi.org/10.1155/2012/561406

Copy DOI

Abstract

In the area of medical image analysis, 3D multimodality image registration is an important issue. In the processing of registration, an optimization approach has been applied to estimate the transformation of the reference image and target image. Some local optimization techniques are frequently used, such as the gradient descent method. However, these methods need a good initial value in order to avoid the local resolution. In this paper, we present a new improved global optimization approach named hybrid particle swarm optimization (HPSO) for medical image registration, which includes two concepts of genetic algorithms—subpopulation and crossover.

Highlights

In the area of medical image analysis, multimodality 3D image registration is an important issue [1]
We perform several registration experiments with both test functions and medical volume data to evaluate the performance of the proposed hybrid particle swarm optimization (HPSO) technique
This paper introduces a new global optimization approach named hybrid particle swarm optimization (HPSO) which incorporates two concepts: subpopulation and crossover of genetic algorithms into the conventional Particle swarm optimization (PSO)

Summary

Introduction

In the area of medical image analysis, multimodality 3D image registration is an important issue [1]. We estimate the parameter of transformation by optimizing a cost function (similarity metric) in the processing of registrations. To overcome the local resolution problem, the genetic algorithm (GA), which is one of the global optimization techniques, has been proposed for medical image registrations [5]. We used an estimation of the transformation that registers the fixed image and moving image by maximizing their cost function (similarity metric) as shown in (1):. Attempting to find the most complex overlapping regions, we should maximize the individual entropies and minimize the joint entropy which could explain each other well

Hybrid Particle Swarm Optimization

Experimental Results

Conclusion and Future Works