Validating Resting State fMRI Derived Brain Connectivity

描述（由申请人提供）：该项目名为“验证静息状态fMRI衍生的大脑连接”，代表了一项大规模的努力，以严格建立静息状态fMRI（fcMRI）衍生的大脑连接测量的可靠性。这种研究大脑区域之间连接的空间模式的方法最近出现了一个密集的增长期。最近的出版物表明，该方法可能能够深入了解大脑区域之间相互作用的大规模结构，这些区域支持大脑在人类健康中的综合功能。还有越来越多的文献表明，偏离正常观察到的连接模式可能是许多精神健康疾病（包括精神分裂症，抑郁症和自闭症）的基本致病因素。在应用这种方法时，在计算连通性的统计测量之前，必须对数据应用大量的处理步骤。目前缺乏知识的影响，不同的预处理方法对使用fcMRI获得的结果的可靠性。所需的一种类型的处理涉及去除由大脑中的心脏和呼吸引起的脉动引起的信号中的变化。目前通常采用简单的数字滤波方法。理论上的考虑表明这种类型的处理可能不会给出最佳结果。本研究将研究通过使用更复杂的方法可能实现的改善，这些方法涉及利用在MRI扫描时从受试者测量的心脏和呼吸周期记录。影响结果的准确性和可靠性的另一个因素涉及为了可以进行受试者之间的比较而必须执行的处理。为了在不同受试者的相应大脑区域之间进行比较，必须使大脑扫描对齐。在神经科学界众所周知，用于实现这一目标的不同数学工具是不完美的，不同的方法会引入不同的误差。本研究将调查的差异，从不同的方法来实现这一对齐的应用程序的可靠性。为了使研究结果具有足够的普遍性，以作为研究者的指南，我们将从大量健康个体中获取fcMRI数据，并使用称为bootstrap设计的研究设计。将从大约100名志愿者中采集多个MRI数据集。所有受试者的数据将采用研究中考虑的每种不同处理方法进行处理。该方法遵循选择受试者子集和计算fcMRI测量值的程序。然后，根据上述结果计算受试者内和受试者间的信度。这个基本过程被重复大量的时间，建立一个大的测量集，从其中可以计算出可靠的统计数据。本研究的结果将有助于fcMRI领域的研究人员进行更明智的研究设计，并有助于解释结果的有效性。这项研究代表了验证fcMRI作为表征正常和病理脑功能的生物标志物的重要下一步。这项研究旨在验证一种新兴的方法，即静息状态功能磁共振成像（fmri）在理解人类和动物大脑连接方面的结果。它将有助于指导从业人员了解利用这一方法取得最佳结果的技术要求。这项拟议的研究对基础神经科学和临床神经科学都做出了贡献，并将为未来研究正常大脑连接性和不同患者群体连接性中断的特征提供基础。

DESCRIPTION (provided by applicant): This project titled 'Validating resting state fMRI derived brain connectivity' represents a large scale effort to establish rigorously the reliability of brain connectivity measures derived from resting state fMRI (fcMRI). This method of studying the spatial pattern of connectivity between regions of the brain has recently seen an intensive period of growth. Recent publications indicate the method may be able to give insight into the large scale structure of interactions between brain regions that support the integrated functioning of the brain in human health. There is also a growing body of literature that indicates deviations from the normally observed pattern of connectivity may be a fundamental causative factor in many mental health disorders including schizophrenia, depression, and autism. In applying this method there are a large number of processing steps that must be applied to the data before the statistical measures of connectivity are calculated. There is currently a lack of knowledge as to the effects of different methods of preprocessing on the reliability of the results obtained using fcMRI. One type of processing required involves the removal of variations in the signal resulting from cardiac and respiratory induced pulsations in the brain. Currently simple digital filtering methods are usually used. Theoretical considerations indicate this type of processing may not give optimal results. This study will investigate the improvements that may be achieved by using more complex methods that involve utilizing recordings of the cardiac and respiratory cycle measured from the subject at the time of MRI scanning. Another factor that influences the accuracy and reliability of the results concerns processing that must be performed in order that comparisons between subjects may be made. In order that comparisons may be made between corresponding brain regions of different subjects the brain scans must be brought into alignment. It is well known in the neuroscience community that the different mathematical tools used to accomplish this are imperfect and that different methods introduce different errors. This study will investigate the differences in reliability that result from the application of the different methods to accomplish this alignment. In order that the results of the study will be sufficiently general to serve as a guide for investigators we will acquire fcMRI data from a large number of healthy individuals and use a study design known as a bootstrap design. Multiple MRI data sets will be acquired from each of approximately 100 volunteers. The data from all subjects will be processed with each of the different processing methods considered in the study. The methodology follows a procedure of choosing subsets of the subjects and calculating fcMRI measures. Next measures of the intra subject and intersubject reliability are calculated from the proceeding results. This basic procedure is repeated a large number of times building a large set of measurements from which reliable statistics may be calculated. The results derived from this study will serve to enable researchers in the field of fcMRI to conduct more informed study design and aid in the interpretation of the validity of results. This study represents an essential next step in the validation of fcMRI as a biomarker for the characterization of normal and pathological brain function. The research is intended to validate the results of an emerging method known as resting state fmri in understanding human and animal brain connectivity. It would serve to guide practitioners of the technical requirements for achieving optimal results utilizing this method. This proposed research contributes to both basic neuroscience and clinical neuroscience and will provide the foundation for future studies characterizing brain connectivity in normals and disruptions of connectivity in various patient populations.