Automatic Brain Morphometry and Volumetry Using SPM on Cognitively Impaired Patients

Alzheimer disease (AD) is the most common type of dementia and mild cognitive impairment (MCI) is a syndrome with an increased risk of developing dementia. As neuroimaging is relevant in their diagnosis, the purpose of this paper is to develop an automatic classification methodology of AD, MCI and control patients. A total of 83 subjects provided by ADNI (Alzheimer's Disease Neuroimaging Initiative) were studied (26 controls, 24 MCI and 33 AD patients) to develop an automatic method. It allows voxel-based morphometry (VBM) optimized by Dartel coregistration and complemented with a volumetric quantification subalgorithm. The developed algorithm implements VBM automatically and segments the three main brain tissues accurately. Difference between controls and AD increased during follow-up (Anova, p = 0,001). Finally statistical parametric maps were created from output images, where morphometric and volumetric differences can be appreciated. This algorithm automates brain imaging volumetry in cognitive impaired patients and control subjects. Even with the described limitations, the developed methodology is fast and user independent and it improves the traditional voxel-based morphometry algorithm.