Quantification of Alzheimer’s Disease Brain Tissue Volume by an Enhanced Possibilistic Clustering Technique Based on Bias-Corrected Fuzzy Initialization

Abstract

Alzheimer's disease weakens the volume of brain tissues such as Gray Matter (GM), white Matter (WM) and Cerebrospinal Fluid (CSF). Precise quantification of tissue volume helps to diagnose this dementia. Although the Fuzzy C-Means (FCM) algorithm and its variants have been widely exploited for measurement of brain tissues volume, its memberships do not always correspond well to the membership degrees of the voxels. To improve this weakness, we propose, an enhanced fuzzy segmentation approach based on the Possibilistic C-Means (PCM) algorithm for deriving fuzzy tissue maps, and the Bias-Corrected FCM algorithm to get the centers initial partition. The hybrid clustering approach exploited for GM, WM and CSF volumes quantification using anatomical and functional brain images extracted from a real database of Alzheimer disease patients. For the purpose of comparison, the conventional FCM and PCM algorithms are executed. We demonstrate that the proposed approach is more robust to noise and partial volume effects than conventional fuzzy algorithms.