Evaluation of Ischemic Stroke Region From CT/MR Images Using Hybrid Image Processing Techniques

Evaluation of Ischemic Stroke Region From CT/MR Images Using Hybrid Image Processing Techniques

Rajinikanth V. (St. Joseph's College of Engineering, India), Suresh Chandra Satapathy (Kalinga Institute of Industrial Technology (Deemed), India), Nilanjan Dey (Techno India College of Technology, India) and Hong Lin (University of Houston – Downtown, USA)
Copyright: © 2018 |Pages: 26
DOI: 10.4018/978-1-5225-5246-8.ch007

Abstract

An ischemic stroke (IS) naturally originates with rapid onset neurological shortfall, which can be verified by analyzing the internal regions of brain. Computed tomography (CT) and magnetic resonance image (MRI) are the commonly used non-invasive medical examination techniques used to record the brain abnormalities for clinical study. In order to have a pre-opinion regarding the brain abnormality in clinical level, it is essential to use a suitable image processing tool to appraise the digital CT/MR images. In this chapter, a hybrid image processing technique based on the social group optimization assisted Tsallis entropy and watershed segmentation (WS) is proposed to examine ischemic stroke region from digital CT/MR images. For the experimental study, the digital CT/MRI datasets like Radiopedia, BRATS-2013, and ISLES-2015 are considered. Experimental result of this study confirms that, proposed hybrid approach offers superior results on the considered image datasets.
Chapter Preview
Top

Introduction

Brain is the most important internal organ, responsible to control the entire operation in human body. The major illness in the brain includes the stroke and the tumor. In order to plan for the proper treatment process, these sickness to be evaluated using an appropriate clinical level diagnose procedures. During the brain condition evaluation procedure a dedicated hardware systems are adopted to record the vital brain information based on the brain signals and brain images. From previous studies, it can be noted that, imaging procedures will provide clear information regarding the brain abnormality compared with the signal assisted evaluation (Palani & Parvathavarthini, 2017). The present chapter considers the evaluation of the brain images recorded using the Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) approaches.

The brain strokes are the destructive illness in human community with enlarged morbidity and mortality rates. The current declaration by the World Stroke Organization (WSO) authenticates that; brain stroke is the second primary cause for disability in human community. The recent yearly statement of WSO also indicates that, each year 17 million persons suffer due to stroke worldwide, among them approximately 6 million die due to severe stroke and remaining 5 million persons experience everlasting disability. Due to its importance, brain stroke examination emerged as the significant research domain among the researchers.

Usually, a brain stroke is caused because of neurological deficit and also the interruption of blood supply due to malfunction in brain blood vessels which will lessen the oxygen and the nutrient supply to the brain tissue (Palani et al., 2016). In order to plan for the treatment process, it is essential to know the reason and location of the stroke. After getting the cause and region of stroke, the doctors can initiate and execute feasible treatment procedure to heal the patient.

An Ischemic Stroke (IS) naturally originates with rapid onset neurological shortfall, which can be verified by analyzing the internal regions of brain. CT and MRI are the commonly used non-invasive medical examination techniques widely considered to record the brain abnormality during the clinical study. In order to identify the disease severity and also to plan for the further treatment process, it is essential to analyze the brain images with the help of medical experts. In recent years, most of the imaging centers have the capability to record digital CT/MR images of the brain. In order to have a pre-opinion regarding the brain abnormality in clinical level, it is essential to use a suitable image processing tool to appraise the digital CT/MRI.

A considerable number of image processing approaches are adopted by the researchers to examine the CT and MRI (Bauer et al., 2013; Maksoud et al., 2015). In this chapter, a hybrid approach based on the Social Group Optimization (SGO) assisted Tsallis entropy function and Watershed (WS) algorithm is proposed to examine the considered test images. In which, the SGO + Tsallis entropy technique is chosen as the pre-processing practice, consider to enhance the brain image and the watershed segmentation approach is adopted as the post-processing procedure to extract the stroke lesion. In this chapter, a comparative examination is presented among Watershed (WS) algorithm, Active Contour (AC) segmentation and Markov Random Field (MRF) procedures existing in the literature (Rajinikanth et al., 2017). The experimental work is implemented using the Matlab software and the result of this study confirms that, WS based segmentation helps to achieve better values of image similarity and statistical measures compared with the AC and MRF based approaches.

The main motivation behind the proposed hybrid image processing approach is to develop a soft-computing assisted image processing tool to examine both the CT/MRI images with improved accuracy. In real time image processing applications, it is necessary to have a unique image processing tool; which can assist the doctors to examine the brain images recorded with CT and MRI.

Complete Chapter List

Search this Book:
Reset