The Alzheimer's Disease Neuroimaging Initiative (ADNI) is a large multi-site study in which serial clinical, biological, neuropsychological and neuroimaging data are being collected from 200 healthy controls, 400 individuals with mild cognitive impairment (MCI) and 200 patients with mild Alzheimer's disease (AD). This proposed ancillary study aims to analyze all ADNI structural and metabolic neuroimaging data to characterize morphometric and metabolic changes in early AD, with the goal of determining optimal predictor variables for identifying individuals at risk for developing progressive AD-related neurodegeneration. The variables thus identified could serve as surrogate endpoints in Phase 2 and 3 clinical treatment trials. To achieve these goals, MRI, PET, and cognitive data will be downloaded from the publicly accessible ADNI database. Methods based on FreeSurfer software will be used to perform automated volumetric segmentation, cortical surface reconstruction and cerebral parcellation on all baseline structural MRIs to obtain measures of regional cortical thickness and subcortical volumes. Automated, longitudinal, within-subject change analyses will be performed on serial MRIs to determine region-specific structural change trajectories related to disease progression. Semi-automated procedures for quantifying metabolic activity within MRI-derived anatomically- defined regions of interest will be applied to PET data from the baseline session to determine effect size of disease-related differences in metabolic activity in all cortical and subcortical structures, before and after correcting for morphometric differences. Within-subject change in metabolic activity over time, before and after correcting for morphometric changes, will be computed to determine region-specific trajectories of disease-related metabolic changes. Multivariate classification analyses will be applied to MRI measures to determine sensitivity and specificity for discriminating controls from subjects with MCI, and for discriminating MCI subjects who convert to AD from those who remain stable in diagnosis. PET and cognitive measures will be added to determine whether they improve classification accuracy. Multivariate analyses will be performed on structural measures obtained from control and MCI subjects during the test sessions of the first study year to determine the optimal set of measures for predicting risk of conversion to AD. Metabolic and cognitive measures will be assessed to determine whether they improve predictive ability. All derived data values and processed image volumes from this study will be made publicly available through the Biomedical Informatics Research Network. This study will significantly enhance understanding of brain changes that occur in early AD;identify candidate neuroimaging biomarkers for use in clinical trials;facilitate the research of other AD investigators;and provide normative data for use in investigation of other aging-related disorders.
The results of this project will provide important new information about the changes in brain structure and metabolism that occur in the earliest stages of Alzheimer's Disease (AD), and relate these measures to change in cognitive performance. This knowledge may improve our ability to predict who is most likely to develop AD and will provide researchers with objective measures that can be used to assess the ability of new treatments to prevent or delay the neurodegeneration associated with AD. Additionally, the high-throughput neuroimaging analysis methods developed here could be used in future studies for detecting and monitoring brain changes that occur in other neurological disorders.
Showing the most recent 10 out of 65 publications
