Integrative Brain Network-Based Analysis for Heterogeneous and Multimodal Neuroimaging Data

点击翻译按钮获取中文摘要

PROJECT SUMMARY This proposal develops state of the art approaches for addressing challenging questions related to the neurobiological mechanisms affecting clinical outcomes of interest in the presence of heterogeneity represented by underlying disease sub-categories and variability in symptoms and other relevant variables across individuals. We focus on developing integrative approaches for brain connectome based analyses, which combines the multi- modal imaging (e.g. fMRI and diffusion MRI) of brain function and structure, clinical and behavioral measures, while accounting for heterogeneity across samples. Our goals involve important questions in neuroscience which have received limited or no attention so far, such as estimating dynamic brain connectivity while incorporating brain anatomical structure, and subsequently examining which dynamic functional connections drive the clinical outcome, accounting for heterogeneity in terms of disease sub-categories when predicting the clinical outcome based on brain measurements which lie on an underlying brain network, and investigating differences in shapes of white matter fiber bundles which drive the clinical outcome. To address such challenging goals, we develop state-of-the-art statistical approaches which incorporate significant innovations and rely on multi-modal neuroimaging data and uses biologically informed priors which yield meaningful solutions. The motivating dataset is the Grady Trauma Project, which contains neuroimaging, behavioral, and clinical data on subjects who were exposed to trauma and developed some degree of PTSD. We will test our approaches on an external PTSD validation dataset from the ENIGMA-PTSD-PGC consortium. Our methodology development will include proposing novel approaches for (a) the joint modeling of multiple graphical models using network-valued regression; (b) using brain anatomical knowledge to inform the estimation of dynamic connectivity and subsequently using the dynamic functional connections to predict the clinical outcome of interest; (c) developing novel approaches for the joint estimation of multiple regression models corresponding to varying subgroups while incorporating network information characterizing the covariates, and (d) developing Bayesian approaches for 3- dimensional shape estimation for fiber tracts in the brain using anatomically informed priors, and subsequently using the shapes of the estimated fiber bundles to predict the clinical outcomes of interest. We also develop a robust strategy for the validation of the proposed methods and we also provide an outline for developing software and sharing them openly with researchers and interested parties. This application addresses several clinical significant questions in neuroimaging research which have not been explored before due to the lack of state of the art statistical methodology, and is expected to make important methodological, scientific, clinical and translational contributions. .