Evaluation of symmetry plane using genetic algorithm

Abstract

Normal human brain is not perfectly symmetrical even it has a high degree of bilateral symmetry. Identification of plane of symmetry is a critical step in brain image analysis. Brain symmetry gives two hemispheres. This paper proposes a new method for evaluating best symmetry plane of brain images using Genetic Algorithm (GA) from Magnetic Resonance Images (MRI). MRI scan is the best modality for brain images, which helps to look for tumors and other diseases in the brain. Symmetry can be reflected in MR image representing axial/horizontal and coronal slices of the brain. Genetic algorithm is an optimization technique which helps to find the optimized solutions/feasible solutions. GA adopted to achieve better results, faster processing times and it is used in more applications. The GA optimization is performed using the selection, mutation and crossover. Major operations are preprocessing, random selection of coordinates and genetic algorithm. Morphological operators help to extract the region of interest from brain image and unwanted portions of MRI images are removed. This proposed method provides better symmetry value with an accuracy of 90%. This performance leads to good approximation of brain tumor segmentation.