Identification of diffusion, kurtosis, and propagator MRI markers of Alzheimer’s disease pathology in post-mortem human tissue

Abstract

Abstract Alzheimer’s disease (AD) is an irreversible degenerative brain disease affecting 6.7 million Americans and while the hallmark AD pathologies of plaques and tangles follow a stereotyped progression during the course of the disease, clinical markers for early diagnosis are lacking and approximately 20% of all AD cases are ultimately misdiagnosed. Conventional clinical MRI is capable of reporting severe brain atrophy, but fails to recognize earlier biomarkers associated with more subtle cellular and molecular changes. Microstructural Magnetic Resonance Imaging (MRI) techniques are promising to address this challenge and may sensitively detect and distinguish tissue degeneration, tauopathies, and beta amyloid plaques to improve accuracy of diagnosis and enable early detection. The objective of this study was to identify and compare the most promising microstructural markers of AD pathology over a range of diffusion and relaxometry-based MRI techniques from conventional to advanced. To accomplish this, we performed MRI microscopy of post-mortem human temporal lobe specimens (n = 14) at high resolution and image quality and evaluated the relative influence of metrics across multiple microstructural MRI frameworks using principal component analysis (PCA). We performed additional correlation analysis between metrics identified by PCA and clinical neuropathology scores of Braak stage and plaque and tangle load. Hippocampal diffusion and restriction metrics contributed most to the first principal component, and the correlation with Braak score was positive for diffusivity and negative for restriction metrics. Additionally, the MAP-MRI propagator anisotropy (PA) metric of microscale anisotropy was strongly and negatively correlated with AD pathology while the conventional fractional anisotropy (FA) metric showed little or no correspondence and there was not a strong association between FA and PA by PCA. Entorhinal cortex findings were minimal except for reported increases in restriction due to plaque content. Taken together, our findings suggest that microstructural MRI metrics of restriction and diffusion are most prominent and may reflect degenerative processes in AD brain tissue and that microscale anisotropy may be more advantageous than conventional FA for the detection of subtle and earlier cellular changes in AD.