Early and differential diagnosis of Alzheimer's disease

This thesis focuses on early and differential diagnosis of Alzheimer’s disease (AD) using multimodality imaging, cognitive measures and artificial intelligence methods of classifications. The findings suggest that measures of semantic free recall and immediate learning are better at detecting preclinical and prodromal cases of Alzheimer’s disease (AD). Similarly, neuronal loss in the ventral tegmental area (VTA) appears a very early occurrence in AD, influences hippocampal volume loss and memory performance and can predict future cognitive decline in ageing. Overall the VTA appears to be a better proxy of disease than amyloidosis, since this latter is not associated with clinical symptoms but it is strongly associated with age. In addition, measures of region to region functional connectivity in temporal cortex were found to be better at distinguishing healthy cases from cases of prodromal AD than measures of impaired activity in the posterior cingulate cortex. In fact, variance in activity in this region appeared to be associated largely with age rather than explained by disease severity as expressed by hippocampal volume.