Computer Aided Diagnosis of Alzheimer’s Disease Using Principal Component Analysis and Bayesian Classifiers

Abstract

Functional brain imaging with PET (Positron Emission Tomography) and SPECT (Single Photon Emission Computed Tomography) has a definitive and well established role in the investigation of a variety of conditions such as Alzheimer’s Disease (AD). Nowadays the inspection of PET and SPECT images is performed by expert clinicians, but usually entails time consuming and subjective steps. This work aims at providing an automatic tool to assist the interpretation of SPECT and PET images for the diagnosis of AD. The main problem to be handled is the so-called small size sample, which consists in having a small number of available images compared to the large number of features. This problem is faced up by reducing intensively the dimension of the feature space by means of Principal Component Analysis (PCA). Our approach is based on bayesian classifiers, which uses the a posteriori information to determine to which class the subject belongs, yielding 88.6% and 98.3% accuracy for SPECT and PET images respectively.