Alzheimer’s disease is a neurodegenerative disease and the most common cause of dementia. Apart from an early and accurate diagnosis, the ability to track progressive changes is important for the development of disease modifying treatments. Diffusion magnetic resonance imaging is a potential method to detect the microstructural changes in the gray matter, which appear years prior to cortical atrophy and clinical symptoms.
 In this thesis, multi-tissue constrained spherical deconvolution is used to model three main tissue classes in the brain (gray matter, white matter and cerebral spinal fluid) based on the diffusion signal. By comparing the tissue fractions in healthy elderly with patients in different stages of the Alzheimer’s disease spectrum using gray matter based spatial statistics, this work demonstrates the potential of the tissue fractions to investigate microstructural changes. The gray matter fraction was lower in patients on the AD spectrum, while the cerebral spinal fluid fraction was higher. These differences are in line with the expected results and appear in pathological affected areas. Compared to another diffusion based metric (mean diffusivity) the tissue fractions showed a higher sensitivity and detected differences in an earlier stage. Overall, this indicates the high potential of these metrics to detect early microstructural changes in Alzheimer’s disease.