Cluster analysis and decision trees of mr imaging in patients suffering alzheimer’s

Abstract

The use of novel analytical techniques (such as data clustering and decision trees) that can model and predict patient disease outcomes has great potential for assessing disease process and progression in Alzheimer’s disease and mild cognitive impairment. For this study, 43 different variables (generated from image data, demographics and clinical data) have been compiled and analyzed using a modified clustering algorithm. Our aim was to determine the influence of these variables on the incidence of Alzheimer’s and mild cognitive impairment. Furthermore, we used a decision tree algorithm to model the level of “importance” of variants influencing this decision.