Early Prediction of Alzheimer's Disease Using Longitudinal Volumetric MRI Data from ADNI

• Activity Number: 253
• Type: Contributed
• Date/Time: Monday, August 1, 2016 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Imaging
• Author(s): Yingjie Li* and Liangliang Zhang and Tapabrata Maiti and David C. Zhu and Jongeun Choi
• Companies: Michigan State University and Michigan State University and Michigan State University and Michigan State University and Michigan State University
• Keywords: Alzheimer's disease ; volumetric MRI ; prediction ; FPCA ; ADNI

Abstract

In this paper, we use the longitudinal volumetric MRI of five regions of interest (ROIs) (hippocampus, entorhinal cortex, middle temporal cortex, fusiform gyrus and whole brain) which were known vulnerable to Alzheimer's disease (AD) to predict conversion from mild cognitive impairment (MCI) to AD. To analyze the longitudinal data, a functional principal components analysis(FPCA) technique which is specifically used for sparse functional data called Principal Analysis through Conditional Expectation (PACE) is employed to estimate the functional principal component scores (FPCSs). For each biomarker, we estimate the FPCSs for each subject. Then use logistic regression with FPCSs as regressors for predicting whether the subject will convert to AD after certain years. We compared the accuracy, sensitivity and specificity in prediction based on individual ROIs as well as the combination of them.