Alzheimer's disease (AD) is a neurodegenerative disease characterized by progressive dementia and affecting almost 10 percent of Americans of 65 years or older. The pathological basis of the disease is believed to involve deposition of cerebral amyloid, which can be detected with CSF sampling or PET scanning. However AD progression rates are highly variable and amyloid positivity alone accounts for a moderate increase to the risk of developing AD in the lifetime. Cerebral blood flow (CBF) is a measure of cerebrovascular integrity and is also tightly coupled to regional cerebral metabolism, thus providing a biomarker for functional neurodegeneration. Reductions in regional CBF corresponding to clinical deterioration in prodromal and fully developed AD have previously been reported in cross-sectional studies. However its trajectory from cognitively normal condition to the diseased stage and also its interaction with amyloid deposition is incompletely understood. Arterial Spin Labeled (ASL) perfusion MRI provides cost-effective non-invasive quantification of CBF and can be measured in routine MRI settings, but suffers from low signal to noise ratio necessitating use of advanced signal processing technique to assess physiological changes. In this proposal, we will characterize longitudinal changes in regional CBF in AD patients using data obtained from Alzheimer's disease neuroimaging initiative (ADNI) and assess the predictive value of regional CBF as an early stage AD biomarker. Our focus will be to identify the brain regions showing the earliest and most consistent changes in CBF in AD subjects. The project will involve developing customized image processing pipeline for processing ASL MRI data from the third and fourth stages of ADNI (ADNI2 and ADNI3), which will include data cleaning and developing automated data quality evaluation metric. ADNI3 data has been acquired using multiple ASL acquisition methods in different scanning platforms, and we will perform a systematic study of its effect on resulting CBF maps. Further, data will be homogenized across different ASL acquisition methods and scanning platforms to facilitate joint analysis of data across scanning platforms. Finally, we will also compare regional CBF and regional structural volumetry to test the hypothesis that they provide complementary information for disease status and progression as well as to assess their combined ability towards the same.
This proposal will characterize the longitudinal changes in regional cerebral blood flow, which is sensitive to Alzheimer's neurodegeneration and can be measured noninvasively using arterial spin labeled (ASL) perfusion MRI. This project will develop procedures for analyzing longitudinal ASL MRI data obtained in the Alzheimer's disease neuroimaging initiative (ADNI) study, and will evaluate the utility of ASL MRI as preclinical or prodromal biomarker of Alzheimer's disease.