Alzheimer's disease (AD) is a neurodegenerative disease characterized by progressive dementia. Diffusion tensor imaging (DTI) has been used widely to delineate the white matter integrity/degeneration in AD. Automated fiber quantification (AFQ) method is a fully automated method, which can identify the major white matter fiber tracts and then evaluate the white matter properties. The main purpose of this study is to assess the white matter integrity and abnormities in a cohort of amnestic Mild Cognitive Impairment (aMCI) and AD patients as well as normal controls (NCs). We tested the utility of diffusion tensor imaging measures along white matter tracts as features to classify the AD from NCs using support vector machine. The results showed that we have detected credible and robust potential early biomarker that may be useful for clinical application in AD.