Automated Diagnosis of Alzheimer´s Disease with Degenerate SVM-Based Adaboost

Alzheimer disease (AD) is known as the most common form of dementia, which imposes a considerable burden on society. In this paper, we focus on the automated diagnosis of Alzheimer disease. Based on the researches on neuropathology, we adopt the thickness of cortex regions from the magnetic resonance imaging (MRI) to characterize the pathology of AD. 3D reconstruction technique is utilized to extract feature vectors from the structured MRI data. To improve the classification quality of our method, we proposed a new classification method which is Based on the combination of SVM and Adaboost. Experiment results show that our method performs well, and can reaches higher classification accuracy than classical classification methods such as k-Nearest Neighbor (KNN), Linear Discriminant Analysis (LDA), Support Vector Machine (SVM), and Gaussian mixture model (GMM).