Application of Pixel Intensity Based Medical Image Segmentation Using NSGA II Based Opti MUSIG Activation Function

Abstract

Medical image segmentation is a challenging task for analyzing the magnetic resonance (MRI) images. These type of images contain missing or diffuse organ/tissue boundaries due to poor image contrast. Medical image segmentation can be addressed effectively by genetic algorithms (GAs). In this article, an application of pixel intensity based medical image segmentation is presented by the non-dominated sorting genetic algorithm-II (NSGA II) based optimized MUSIG (Opti MUSIG) activation function with a multilayer self organizing neural network (MLSONN) architecture. This method is compared with the process of medical image segmentation by the MUSIG activation function with the MLSONN architecture. Both the methods are applied on two real life MRI images. The comparison shows that NSGA II based Opti MUSIG activation function performs better medical image segmentation than the MUSIG activation function based method.