Classification of dementia from FDG-PET parametric images using data mining

Abstract

It remains a challenge to identify the different types of dementia and separate these from various subtypes from the normal effects of ageing. In this paper the potential of parametric images from FDG-PET studies to aid the classification of dementia using data mining techniques was investigated. Scalar, joint, histogram and voxel-level features were used in the investigation with principal component analysis (PCA) for dimensionality reduction. The logistic regression model and the additive logistic regression model were applied in the classification. The results show that cerebral metabolic rate of glucose consumption (CMRGlc) was efficient in the classification of dementia and data mining using voxel-level features with PCA and the logistic regression model method achieving the best classification.