Fuzzy-based segmentation of brain parenchymal regions with Alzheimer's disease into cerebral cortex and white matter in 3.0-T magnetic resonance images

It would be very important to estimate the degree of cerebral atrophy based on cortical regions for diagnosis of Alzheimer's disease (AD). However, it would be still challenging to segment brain parenchymal regions with AD into cerebral cortex and white matter when the boundary between them is unclear due to the presence of AD showing in magnetic resonance (MR) images. Our purpose of this study was to develop an automated segmentation of the brain parenchyma into cerebral cortical and white matter regions with AD in three-dimensional (3D) T1-weighted MR images. Our proposed method consisted of extraction of a brain parenchymal region based on a brain model matching and segmentation of the brain parenchyma into cerebral cortical and white matter regions based on a fuzzy c-means (FCM) algorithm. We applied the proposed method to MR images of the whole brain obtained from 9 cases, including 4 AD cases and 5 control cases. The mean volume percentages of the brain parenchymal region in the respective AD patients and controls were 41.7% and 45.2% for cortical cortex region, 58.3% and 54.8% for white matter region, respectively.