Periodic Reporting for period 1 - uImaging (MRI based mapping of microscopic brain composition in a mouse model of Alzheimer’s disease)
Reporting period: 2020-05-01 to 2022-04-30
This project aims to establish a neuroimaging framework which informs about microscopic brain composition and apply it to study neurodegeneration in a genetic mouse model of Alzheimer’s disease (AD).
AD is the most common cause of dementia, characterized by progressive neurodegenerative changes leading to a gradual cognitive loss and memory impairment. Pathological changes in AD are characterized by the accumulation of proteins to form amyloid plaques and tau-tangles as well as by variations in microscopic brain composition such as cell body density, neurite density, axon and myelin morphology, which can precede macroscopic brain atrophy and clinical symptoms even by decades. Thus, developing imaging biomarkers which can inform about these changes throughout the brain is a crucial step for early diagnosis and a better understanding of the disease.
To achieve this goal, this proposal takes an interdisciplinary approach combining state-of-the-art magnetic resonance imaging (MRI) modalities and novel computational methods, and has the following objectives:
1. Develop and validate a neuroimaging framework which encompasses advanced diffusion MRI and multi-exponential T2 MRI to map microscopic brain composition (markers of soma size and density, neurite density, axon diameter and myelin fraction). Pre-clinical imaging will be performed both in-vivo and ex-vivo in wild type mice and compared with histological staining.
2. Apply this methodology to investigate the changes in microscopic brain composition with the progression of AD in the APP/PS1 transgenic mouse model which develops Aβ-plaques, mimicking the early disease stages.
3. Study the association between neurodegeneration and the development of Aβ-plaques, that will be quantified based on both MR micro-imaging with ultra-high resolution and histological staining.
Thus, this proposal has both basic science and applied components and will serve as a stepping stone for Neuroimaging of AD.
AD is the most common cause of dementia, characterized by progressive neurodegenerative changes leading to a gradual cognitive loss and memory impairment. Pathological changes in AD are characterized by the accumulation of proteins to form amyloid plaques and tau-tangles as well as by variations in microscopic brain composition such as cell body density, neurite density, axon and myelin morphology, which can precede macroscopic brain atrophy and clinical symptoms even by decades. Thus, developing imaging biomarkers which can inform about these changes throughout the brain is a crucial step for early diagnosis and a better understanding of the disease.
To achieve this goal, this proposal takes an interdisciplinary approach combining state-of-the-art magnetic resonance imaging (MRI) modalities and novel computational methods, and has the following objectives:
1. Develop and validate a neuroimaging framework which encompasses advanced diffusion MRI and multi-exponential T2 MRI to map microscopic brain composition (markers of soma size and density, neurite density, axon diameter and myelin fraction). Pre-clinical imaging will be performed both in-vivo and ex-vivo in wild type mice and compared with histological staining.
2. Apply this methodology to investigate the changes in microscopic brain composition with the progression of AD in the APP/PS1 transgenic mouse model which develops Aβ-plaques, mimicking the early disease stages.
3. Study the association between neurodegeneration and the development of Aβ-plaques, that will be quantified based on both MR micro-imaging with ultra-high resolution and histological staining.
Thus, this proposal has both basic science and applied components and will serve as a stepping stone for Neuroimaging of AD.
Over the first 5 months of this fellowship I have performed the following tasks:
1. I have finalized my work on theoretical and computational methods which employed simulations of diffusion MRI in realistic neuronal configurations to study the effect of soma size and dendritic branching on the dMRI. These simulations allow us to better understand the imaging contrast and choose acquisition parameters which maximize the effects of soma size and density (https://arxiv.org/abs/2009.11778(opens in new window))
2. I have worked on a pipeline which employs a machine learning approach based on Random Forest regressions to estimate the parameters of multi-compartment relaxometry models and to map the myelin water fraction.
3. I have started optimizing the in-vivo diffusion acquisition in control animals.
4. I have established the colony of transgenic Alzheimer’s mice.
1. I have finalized my work on theoretical and computational methods which employed simulations of diffusion MRI in realistic neuronal configurations to study the effect of soma size and dendritic branching on the dMRI. These simulations allow us to better understand the imaging contrast and choose acquisition parameters which maximize the effects of soma size and density (https://arxiv.org/abs/2009.11778(opens in new window))
2. I have worked on a pipeline which employs a machine learning approach based on Random Forest regressions to estimate the parameters of multi-compartment relaxometry models and to map the myelin water fraction.
3. I have started optimizing the in-vivo diffusion acquisition in control animals.
4. I have established the colony of transgenic Alzheimer’s mice.
1. I have shown for the first time that adanced diffusion MRI approaches such as double diffusion encoding can provide contrast to the dentritic tree branching in grey matter, a property which is not detectable with standard dMRI acquisitions.
2. I have shown in simulatations that Random Forests can be applied to estimate Myelin Water Fractions from T2* measurements in a computational time which is orders of magnitude lower compared to standard non-linear otimisation techniques.
3. I have acquired the first maps of soma density and size for in-vivo mouse brains.
These first steps are crucial for establishing a reliable and sensitive imaging technique which can inform about the microscopic brain composition and then be used to study changes induced by Alzheimer’s disease.
2. I have shown in simulatations that Random Forests can be applied to estimate Myelin Water Fractions from T2* measurements in a computational time which is orders of magnitude lower compared to standard non-linear otimisation techniques.
3. I have acquired the first maps of soma density and size for in-vivo mouse brains.
These first steps are crucial for establishing a reliable and sensitive imaging technique which can inform about the microscopic brain composition and then be used to study changes induced by Alzheimer’s disease.