Characterization of Human Hippocampus using Textural Analysis in Epileptics patient

Abstract

This study is an attempt to study the hippocampus (body, head, tail and sagittal section) in MRI images using computerized textural analysis and hence the main objectives of this study was to characterize the hippocampal tissues into two classes normal and epileptic using textural analysis. The texture were extracted from spatial gray level dependence matrix using a window of 20×20 pixels of angle zero and distance equal one pixel. The images were collected from MRI brain scans for 18 patients represent the classes of the study in the period from 7/2011 to 2/2012. The images were scored by an expert radiologist and the scoring was accepted in case of agreement with findings of EEG. Then the features were extracted from the selected sub-images that show only the region of interest. The classification technique were adopted as a method of pattern identification the images into normal of epileptic class. A linear discriminant analysis using stepwise were used to classify the sample into the predefined classes. The stepwise selected number of features out of fifteen features as the most discriminant features for each hippocampal region; they included: Energy, Entropy, Correlation Inertia, difference average, difference entropy, difference variance, sum variance, um entropy, IDM, Information1 and information2. The result of this study showed that the total classification accuracy was 83.3%, 80.6%, 91.7%, and 79.6% for body, head, tail and sagittal respectively. The sensitivity was 72.2, 72.2, 94.4% and 79.6. The Specificity was 94.4%, 88.92%, 88.9% and 79.6% respectively.