Unraveling Alzheimer's Disease: Relations Between Neuroimaging and Neuropsychological Measures
Chin Fatt, Cherise Regina
MetadataShow full item record
Alzheimer’s Disease (AD) is a multifaceted disease that includes psychiatric, cognitive, and neurologic components. To better understand the progression of AD it is important to investigate how cognition and neuroimaging biomarkers interact at different stages of the disease. This area of research could be enhanced by incorporating a potentially useful neuroimaging technique, namely resting state functional MRI (rsfMRI). In addition to including rsfMRI, researchers are often challenged with selecting the appropriate cognitive test(s) to address their research questions. With multiple areas of cognition being affected at different stages of AD and with many possible cognitive tests to choose from, selecting the appropriate cognitive test can be a diffcult task. The goal of this research was to determine how pathological processes—measured using neuroimaging techniques and multiple cognitive tests—evolve based on the stages of AD. Neuroimaging and cognitive data were acquired from the Alzheimer’s Disease Neuroimaging Initiative database. Data were obtained from participants classified in five groups: (1) 34 participants in the Healthy Control group (HC), (2) 25 participants in the Significant Memory Concern group (SMC), (3) 30 participants in the Early Mild Cognitive Impairment group (EMCI), (4) 31 participants in the Late Mild Cognitive Impairment group (LMCI), and (5) 31 participants in the Early Alzheimer’s Disease group (EAD). The neuroimaging biomarkers of interest were (1) amyloid β (Aβ), (2) regional cerebral glucose metabolism, (3) functional connectivity, (4) cortical atrophy, and (5) subcortical atrophy. The statistical analysis involved two stages: (1) cognitive measures were analyzed to determine the relationships between cognitive measures and participants; (2) neuroimaging biomarkers and cognitive measures were analyzed to determine the relationships between neuroimaging biomarkers, cognitive measures, and participants. The main findings of my study are as follows. First, apart from AD, cognitive performance alone was not a good predictor of the diagnostic group of the participants. Second, adding amyloid status improved accuracy when there were clearer differences within the stages of disease compared to between the stages of disease in the progression of AD. Last, the neuroimaging and cognitive data clustered into four groups, namely, (1) Aβ, (2) cognition and glucose metabolism, (3) cortical thickness, cortical volume, and subcortical volume, and (4) functional brain networks. To conclude, this study reports some important results. The first result was seen in the analysis of the cognitive data where the misclassification of participants to clinical groups occurred mainly when classifying participants in the following groups: in the HC, SMC, EMCI, and LMCI . The misclassification to the MCI group could be due to participants being misdiagnosed and who instead suffer from frontotemporal dementia or psychiatric diseases. Second, for cognition, there were clearer within stage differences compared to between stage differences in the progression of the disease in the analyses involving all participants compared to analyses involving only amyloid positive participants. This finding suggests that once the brain has elevated amyloid deposits, a person will progress through the stages of AD; however the cognitive changes within the different stages of disease will be subtle. Together my dissertation results suggest the many distinguishing brain changes under an AD diagnosis can be assessed using similar neuroimaging techniques and cognitive tests. Overall, the key to an AD diagnosis is elevated Aβ, a psychiatric screening, and assessments of other neurodegenerative disorders which have similar AD symptoms (to reduce misdiagnosis).