Comparative Analysis Across Advanced Diffusion MRI Techniques for Detection of Alzheimer’s Disease Pathology

Abstract

Alzheimer’s disease (AD), an irreversible neurodegenerative brain disease, is characterized by hallmark pathologies such as beta-amyloid plaques, neurofibril tangle, and cellular degeneration, making diagnosis challenging due to comorbidities and other dementias. Traditional MRI can detect late-stage degeneration and morphology changes but fails to sense subtle microstructural changes that happen in the early stages of the disease. This thesis aimed to identify clinically applicable diffusion MRI metrics capable of early detection of AD pathology. Results in this thesis, from a post-mortem microstructural MRI study in the hippocampus and entorhinal cortex using traditional diffusion encodings, highlighted MR metrics that are sensitive to the progression of neurofibril tangle infestation and beta-amyloid plaque presence. While promising diagnostic tools were identified during this study, the clinical implementation of these metrics remains challenging due to the time-intensive acquisitions, MRI hardware needs, and computational demands. In this thesis, these challenges were addressed, with an alternative diffusion acquisition, q-space trajectory imaging (QTI), capable of sensitizing the MRI signal to specific geometries and sizes resulting in reduced acquisition time needs. A validation of the QTI framework was conducted in a series of differing microstructural phantoms and showed key similarities to the traditional diffusion MRI framework metrics along with differing strengths. In an evaluation of both frameworks in AD post-mortem hippocampus, QTI showed improved early detection of AD when compared to the already promising traditional diffusion markers.