Meta-Analysis in Human Brain Mapping

人脑图谱的荟萃分析

• 批准号: 8464784
• 负责人: PETER Thornton FOX
• 金额: $ 60.29万
• 依托单位: UNIVERSITY OF TEXAS HLTH SCIENCE CENTER
• 依托单位国家: 美国
• 财政年份: 2006
• 资助国家: 美国
• 起止时间: 2006-09-15 至 2015-03-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/8464784
• 关键词: Adopted; Adoption; Algorithms; Atlases; Behavioral; Brain; Brain Mapping; Code; Cognitive; Collaborations; Communities; Computational Science; Computer software; Data; Data Set; Databases; Development; Diagnosis; Disease; Fostering; Functional disorder; Funding; Goals; Human; Knowledge Discovery; Label; Laboratories; Literature; Maps; Mental disorders; Meta-Analysis; Methods; Mining; Modeling; Motor; Neurologic; Ontology; Paper; Patients; Pattern; Peer Review; Property; Public Health; Publications; Publishing; Research; Resource Informatics; Resource Sharing; Resources; Rest; Scheme; Sensitivity and Specificity; Software Tools; Speed; System; Taxonomy; Technology; United States National Institutes of Health; base; computerized tools; data mining; data space; developmental disease; improved; independent component analysis; innovation; meetings; morphometry; nervous system disorder; neuroimaging; neuroinformatics; parent grant; public health relevance; rapid growth; relating to nervous system; text searching; tool

DESCRIPTION (provided by applicant): The overall objective of the present proposal is to develop, evaluate, distribute, and apply tools for the BrainMap Project, which provides the human brain mapping (HBM) community with data sets, computational tools, and informatics resources for quantitative meta-analyses and meta-analysis-based data interpretation. The development of coordinate-based, voxel-wise meta-analysis (CVM) has been a breakthrough for HBM, enabling statistically rigorous meta-analysis and systems-level modeling of published results. The HBM literature suitable for CVM meta-analysis was estimated in 2007 to be ~6,000 papers, with ~1000 new conforming papers being published each year. In this renewal, we propose numerous enhancements to widely used coordinate-based meta-analysis tools (Aim 1). Meta-analytic connectivity mapping (MACM) is a new application of CVM, which derives inter-regional connectivity maps from inter-study co-occurrence patterns. In this renewal, we propose to extend the functionality of our MACM tools, construct connectivity atlases through data mining, and validate these atlases by comparison to non-meta-analytic approaches (Aim 2). In BrainMap, behavioral meta-data are coded using a taxonomy developed and progressively refined by the BrainMap team. Recent application of independent component analysis (ICA) to BrainMap extracted intrinsic neural systems that were well, but not perfectly, discriminated by the BrainMap coding scheme. This observation suggests robust computational approaches both for providing a behavioral ontology for intrinsically connected networks and also for programmatic refinement of the BrainMap coding scheme (Aim 3). Finally, it has recently been demonstrated that CVM can be applied to voxel-based morphometry (VBM) structural neuroimaging studies. In Aim 4, we anticipate the rapid growth of VBM research by creating a BrainMap-like database of the standards- compliant VBM literature. These proposed enhancements and extensions of the BrainMap Project will allow this rich set of neuroinformatics tools to continue to meet the research and educational needs for knowledge discovery and data mining in the neuroimaging community. 1

描述（由申请人提供）：本提案的总体目标是为Brainmap项目开发，评估，分发和应用工具，该项目为人脑映射（HBM）社区提供数据集，计算工具和信息学资源，以用于定量的荟萃分析和基于荟萃分析的数据解释。 基于坐标的，体素的元分析（CVM）的开发一直是HBM的突破，从而实现了统计上严格的荟萃分析和已发布结果的系统级建模。 估计适合CVM荟萃分析的HBM文献估计约为6,000篇论文，每年发表约1000篇新的合格论文。 在此续约中，我们提出了广泛使用的基于坐标的荟萃分析工具的大量增强功能（AIM 1）。 荟萃分析连接映射（MACM）是CVM的新应用，它从研究间的共发生模式中得出区域间连接图。 在此续订中，我们建议扩展MACM工具的功能，通过数据挖掘构建连接图像，并通过与非米塔分析方法进行比较来验证这些图书馆（AIM 2）。 在Brainmap中，使用Brainmap团队开发并逐步完善的分类法对行为元数据进行了编码。 独立组件分析（ICA）的最新应用在脑图提取的固有神经系统中，这些神经系统很好，但并非完美地被Brainmap编码方案所歧视。 该观察结果表明，具有强大的计算方法，既可以为本质上连接的网络提供行为本体，又是用于对脑图编码方案的编程改进（AIM 3）。 最后，最近证明CVM可以应用于基于体素的形态计量学（VBM）结构神经影像学研究。 在AIM 4中，我们通过创建符合标准的VBM文献的类似脑形图的数据库来期待VBM研究的快速增长。 这些提议的增强和扩展Brainmap项目将使这组丰富的神经信息学工具能够继续满足神经影像社区中知识发现和数据挖掘的研究和教育需求。 1