Collaborative Research: Bayesian Approaches for Inference on Brain Connectivity

这个合作研究项目将开发新的统计方法来分析和解释脑成像数据。我们需要统计方法来改进信号检测，并对大脑功能模式进行临床相关的洞察。这项研究将促进对大脑不同区域在任务或休息时如何相互作用和共享信息的理解。即将开发的统计方法将有可能影响统计学和神经影像学，并将普遍适用于在受试者组上测量多种类型神经影像学数据的研究。从社会的角度来看，获得的知识将指导临床医生选择最佳的有针对性的治疗，以提高个人的生活质量。该项目将包括对研究生的教育和培训活动。研究结果将被传播到研究界，并用于进一步跨学科的神经影像学合作努力。软件和代码将开发和存储在公共存储库。即将开发的新的统计方法将综合收集到的对不同受试者群体的多种成像方式提供的信息。提出的研究的一个特别重点是表征受试者内部和受试者之间大脑功能的异质性。这项研究将产生灵活的贝叶斯统计方法，可以跨学科共享信息，并考虑到大脑结构和功能机制方面的现有知识。新的综合时空模型将允许在大脑网络中存在高度连接和持久的中枢。动态图形模型方法将增加对功能性大脑连接的动态性质的理解，以及当受试者完成任务时连接是如何中断的。研究人员将把这种新方法应用于患有神经系统疾病（癫痫或精神分裂症）的受试者和作为对照的健康个体的成像数据。了解脑连接组异常在各种神经系统疾病中的作用一直是连接研究的主要焦点。对作为对照的健康个体的数据进行比较分析，将有助于确定不同受试者群体之间连通性的差异，以及它们如何影响多个认知领域。

This collaborative research project will develop new statistical methods for the analysis and interpretation of brain imaging data. Statistical methods that improve signal detection and that lead to clinically relevant insights into the patterns of brain functions are needed. This research will advance understanding of how the different regions of the brain interact and share information with each other during a task or at rest. The statistical methods to be developed will have the potential to impact both statistics and neuroimaging and will apply generally to studies where multiple types of neuroimaging data are measured on groups of subjects. From a societal perspective, the acquired knowledge will guide clinicians in the selection of optimally targeted treatments to improve the quality of life of individuals. The project will include educational and training activities for graduate students. Findings will be disseminated to the research community and used to further interdisciplinary collaborative efforts in neuroimaging. Software and code will be developed and deposited in public repositories.The new statistical methods to be developed will integrate the information provided by multiple imaging modalities collected on groups of subjects. A particular focus of the proposed research is to characterize the heterogeneity of brain functioning both within and between subjects. This research will produce flexible Bayesian statistical methods that can share information across subjects and take into account available knowledge on brain structure and functional mechanisms. New integrative spatio-temporal models will allow for the presence of highly connected and persistent hubs in the brain networks. Dynamic graphical model approaches will increase understanding of the dynamic nature of functional brain connectivity and how connectivity is disrupted when subjects are completing tasks. The investigators will apply the new methods to imaging data from subjects with a neurological disorder (epilepsy or schizophrenia) and data from healthy individuals who will serve as controls. Understanding the role that abnormalities in the brain connectome play in various neurological diseases has been a major focus in connectivity studies. Comparative analyses of data from healthy individuals serving as controls will allow the identification of differences in connectivity across groups of subjects and how they affect multiple cognitive domains.