Neurological image classification for the Alzheimer´s Disease diagnosis using Kernel PCA and Support Vector Machines

An accurate and early diagnosis of the Alzheimer´s Disease (AD) is of fundamental importance for the patients medical treatment. Single Photon Emission Computed Tomography (SPECT) and Positron Emission Tomography (PET) images are commonly used by physicians to assist the diagnosis and rated by visual evaluations, manual reorientations and other time consuming steps. In this work we present a computer assisted diagnosis (CAD) tool for the early diagnosis of the AD, based on Kernel Principal Component Analysis (PCA) dimension reduction of the feature space in combination with Linear Discriminant Analysis (LDA). The extracted information is used to train a kernel-based Support Vector Machine (SVM) classifier, which is able to classify new subjects in an unsupervised manner. This approach outperforms other recently developed multivariate approaches reaching up to 92.31% and 96.67% accuracy values for SPECT and PET images respectively.