Computational modeling of dynamic causal brain circuits underlying cognitive dysfunction in Alzheimer's disease

点击翻译按钮获取中文摘要

Project Abstract Alzheimer’s disease (AD) affects over 5.7 million Americans and is expected to rise to nearly 14 million people by 2060, as the number of people living with this disease doubles every 5 years. AD is a progressive and severely debilitating disease that negatively affects cognitive and memory function and is linked to increased disability in everyday functioning and risk of mortality. Neurodegeneration of focal brain areas in AD progressively impacts large-scale brain circuits, leading to significant cognitive and behavioral impairments. However, little is known regarding aberrant context-dependent dynamic causal interactions between distributed brain regions, and their links to cognitive and memory impairments and neuropathology, across AD clinical stages. Leveraging a productive and high-impact line of research in the current project period, we now propose to address critical gaps in our knowledge of functional circuit mechanisms underlying cognitive dysfunction in AD using innovative computational tools. Our first major goal is to continue to address critical unmet needs in human brain research by developing and validating novel computational tools for identifying context-dependent dynamic causal interactions between distributed brain regions. Building on progress in the current project period, we will further develop novel Multivariate Dynamic Systems Identification-Hamiltonian Monte Carlo techniques taking advantage of recent advances in Bayesian modeling and inferencing. Our computational tools will be validated using optogenetic stimulation with whole-brain fMRI, and stability analysis of normative Human Connectome Project data. Our second major goal is to use MDSI-HMC to investigate aberrancies in dynamic causal circuits underlying cognitive and memory impairment in AD. Our system neuroscience approach will target four key brain systems implicated in AD: default mode network, medial temporal lobe, and two frontal control systems anchored in the frontoparietal and salience networks. To achieve our goals, we will leverage clinical, phenotypic, cognitive, experimental, and state-of-the-art fMRI, and beta amyloid (Aβ) and tau PET, data from multiple NIH-funded AD-specific Human Connectome Projects. Our proposed studies will advance foundational knowledge of cognitive and memory-related circuits across AD clinical stages and their links to neuropathology. More generally, our proposed studies will also contribute novel tools for examining dynamical causal circuits underlying human brain function and dysfunction. The proposed studies are highly relevant to the NIH Focus on AD and PAR- 10-070 which call for innovative characterization of functional brain circuits altered in AD. More broadly, the proposed studies are relevant to the mission of the NIH to encourage development and dissemination of innovative advanced computational tools for clinical neuroscience. We will disseminate our algorithms and software tools to the research community as we have done in the current project period.