Cerebral Cortex Segmentation from MR Brain Images based on Contouring Technique to Detect Alzheimer’s Disease

Abstract

Magnetic Resonance Images (MRI) are extensively used for medical diagnosis to detect various dementia by analyzing the whims in the brain tissues of the human brain. The popularly known dementia is Alzheimer’s Disease (AD), and dissection of brain tissue acts a vibrant role to diagnose AD. One such brain tissue is Cerebral Cortex (CC) or Cortical Thickness (CT), which is majorly used for pathological studies, CC/CT segmentation is the most significant part of diagnosis. This paper reveals the importance of segmenting the cerebral cortex to detect AD. In this method, the segmentation is done in three processes; the first step is to find the cerebral hemispheres, the second step is segmenting the CC in each hemisphere by employing a contour-based approach and in the thirdly, the CC is segmented based on the histogram statistics and intensity profile of the segmented region by contouring technique. Earlier approaches for AD detection are based on brain tissue like white matter and gray matter segmentation. This Proposed method segments CC and yields acceptable results to detect AD in pathophysiological images and the results of this method is evaluated with brain images of IBSR 18 and ADNI using Jaccard (Jac), Dice (Dc) similarity measures, Dissimilarity Jaccard (dJac) and with Sensitivity (Sen), Specificity (Spc) quantitative measures.