功能磁共振动力学模型及脑静态网络研究

【作者】 王昱青；

【导师】 陈华富；

【作者基本信息】 电子科技大学 ， 生物物理学， 2006， 硕士

【摘要】 脑功能磁共振(Functional magnetic resonance imaging ,fMRI)主要依据血氧水平依赖性(blood oxygenation level dependent ,BOLD)对比增强原理进行成像,是目前人们掌握的唯一无侵入、无创伤、可精确定位的研究手段。fMRI具有很高的空间分辨率,而且时间分辨率具有进一步提高的潜力,非常适合神经活动的时空分析和脑的高级功能研究,已受到神经科学、认知科学和临床诊断等领域的极大关注。在fMRI动力学模型研究方面,本文在分析Friston的BOLD动力学模型的基础上,结合Agnes Aubert所建立的电生理-代谢耦合模型,提出了一种扩展的BOLD动力学模型。该模型把脑的电活动和新陈代谢与血动力学模型的血流、血体积联系起来,得到的各参数的理论曲线是合理的。对BOLD模型中输入信号的研究结果表明,不同的输入信号对模型的输出有一定的影响。改进的模型能更好地模拟生理过程。在脑静态网络研究方面,fMRI为研究脑的功能活动和各个脑功能区的相互关系提供了一种有效的手段,不同脑功能区之间的活动方式是网络模式的,它们共同执行许多复杂的脑功能活动。脑静态网络就是这些网络中的一种,它反映脑的本征活动和规律,是脑功能研究的基础。本文采用整合邻域相关分析的分级聚类方法来研究脑的静态网络。在分级聚类方法中,本文引入了一种新的度量各个类之间距离的方法,该方法能够更好地利用功能磁共振数据的时间空间信息。和模型驱动方法相比,新的聚类方法不需要先验信息,比如:刺激模式,它克服了模型驱动的瓶颈。和其它数据驱动方法不同的是:该方法对数据的具体特性没有太多要求,也不存在挑选有用成分的问题,同时也克服了传统聚类方法存在的聚类矩阵计算复杂度高,内存消耗大,计算时间长等缺点。新的聚类方法可以有效地分析静息状态下的脑默认网络。本文通过该方法首次比较了脑静态网络的性别差异。综上所述该种方法对于研究脑静态网络是一个比较有前景的方法。更多还原

【Abstract】 Functional magnetic resonance imaging (fMRI) is mainly based on blood oxygenation level dependent (BOLD). It is the most efficient method that can be used to precisely locate brain activities without invasion. With very high spatial resolution and potential high temporal resolution, fMRI is fit for the spatial and temporal analysis of neural action and the research of advanced brain function. So it is being concerned by many science branches such as neuroscience, cognition and clinical diagnosis et al and is a hot point in current brain research.In the sight of research of fMRI dynamic model, basing on the analysis of Friston’s BOLD model of dynamics, combining Agnes Aubert’s coupling model of brain electrical activity and metabolism, we proposed an extended BOLD dynamic model. The improved model connected the cerebral electrical activity and metabolism with the blood flow or blood volume of the hemodynamic model, and the result of emulator consisted with the real experiment data. The result of researching of input signal indicated that different input signals caused different effect to the output of the model. The improved model could simulate the physiological process better.In the sight of research of cerebral default model, fMRI technique provides some measures for researching of brain functional action and correlation among different brain functional regions. The activity mode of different brain functional regions is a network mode, and these brain functional regions carry out many complex brain functional actions together. The default networks are one of these networks, and reflect the intrinsic activity and rules of brain. It is the base of research of brain function.In this paper, an approach integrating neighborhood local correlation (NLC) and hierarchical clustering analysis (HCA) methods is introduced into investigating default networks. In HC procedure, a new spatial temporal distance measure is proposed for better utilizing spatial-temporal information in fMRI datasets. Comparing to those model-driven methods, it doesn’t need prior information, for instance the stimulus mode,更多还原