Advanced MR Neuroimaging in Early Stage Presenile Dementia

Dementia is a syndrome affecting cognitive functions and behavior, with an overwhelming impact on both patients and caregivers. An estimated number of 35.6 million patients suffers from dementia, with a subset affected before the age of 65 years, i.e. presenile dementia. Establishing the underlying cause of presenile dementia in the early stage proves to be difficult. Morphological brain changes associated with the early stage of dementia are often absent or unspecific on routine clinical magnetic resonance imaging (MRI). Advanced MR neuroimaging techniques may be more sensitive to early neurodegenerative processes as they can capture functional and microstructural brain changes that precede (macro)structural damage. In this thesis, I investigated whether resting state functional MRI (rs-fMRI), arterial spin labeling (ASL), and diffusion tensor imaging (DTI) could serve as early neuropathological markers of presenile dementia.

We found that advanced MR neuroimaging techniques are sensitive to subtle incipient changes in presenile dementia. This facilitates early diagnosis and differentiation of two common causes of presenile dementia – Alzheimer’s disease and frontotemporal dementia (FTD) – and contributes to their respective characterization in a more fundamental way. Additionally, we used ASL, DTI and rs-fMRI in phenocopy frontotemporal dementia, as of yet still described as a clinical syndrome, and showed for the first time brain changes that can be linked to the neurodegenerative process of FTD. Finally, ASL shows value as a clinical imaging tool, but its wider application is hampered by variability between ASL data acquired on MR scanners of different vendors, as each vendor provides their own particular ASL implementation. We assessed to which extent the implementation of similar sequences by two different vendors affect ASL perfusion data in several experimental designs in healthy young controls. The results from these reproducibility studies indicate that we can only compare and exchange quantitative ASL data if implementation is standardized.