带时间约束的Louvain算法在动态脑功能网络模块化中的应用研究#br#

针对在动态脑功能网络的模块化属性研究中,Louvain算法因过度追求模块度值最大化而导致的动态脑功能网络模块辨识度不高的问题,提出了一种带时间约束的Louvain算法。该算法以整个数据采集区间上的模块度值分布为依据构建迭代结束条件,以时间约束来达到模块在规模和数量上的均衡,从而保证模块划分更加合理。将本文算法用于静息态脑功能的模块划分实验时,对比结果显示,与原Louvain算法相比,带时间约束的Louvain算法能够得到更为合理的模块化结果,并可以观测到动态脑功能网络中小规模的模块结构。而采用本文算法用于健康人和自闭症患者的动态脑功能网络模块度对比实验,能够揭示两者在模块化上存在显著差别,从而验证了本文算法的有效性。     

基于功能性核磁共振的汉字语义加工脑机制研究

利用功能性磁共振成像（fMRI）技术探讨文盲和非文盲汉字语义加工脑机制的差异。实验使用汉字语义和图形的判断任务比较了中国人文盲和非文盲在语义加工过程脑机制的差异。结果表明文盲与非文盲汉字语义加工脑机制不同,且非文盲的脑活动更强,具体体现在左侧额中回（BA9）,左侧角回（BA39/40）及左侧颞上回（BA21/22）以及小脑。     

Adaptive integration of local region information to detect fine-scale brain activity patterns

With the rapid development of functional magnetic resonance imaging (fMRI) technology, the spatial resolution of fMRI data is continuously growing. This provides us the possibility to detect the fine-scale patterns of brain activities. The established univariate and multivariate methods to analyze fMRI data mostly focus on detecting the activation blobs without considering the distributed fine-scale patterns within the blobs. To improve the sensitivity of the activation detection, in this paper, multivariate statistical method and univariate statistical method are combined to discover the fine-grained activity patterns. For one voxel in the brain, a local homogenous region is constructed. Then, time courses from the local homogenous region are integrated with multivariate statistical method. Univariate statistical method is finally used to construct the interests of statistic for that voxel. The approach has explicitly taken into account the structures of both activity patterns and existing noise of local brain regions. Therefore, it could highlight the fine-scale activity patterns of the local regions. Experiments with simulated and real fMRI data demonstrate that the proposed method dramatically increases the sensitivity of detection of fine-scale brain activity patterns which contain the subtle information about experimental conditions. Supported by Chair Professors of Changjiang Scholars Program and CAS Hundred Talents Program, National Program on Key Basic Research Projects (Grant No. 2006CB705700), National High-Tech R&D Program of China (Grant No.2006AA04Z216), National Key Technology R&D Program (Grant No. 2006BAH02A25), Joint Research Fund for Overseas Chinese Young Scholars (Grant No.30528027), National Natural Science Foundation of China (Grant Nos.30600151, 30500131 and 60532050), and Natural Science Foundation of Beijing (Grant Nos. 4051002 and 4071003)     

Neurofeedback using functional spectroscopy

Neurofeedback based on real‐time measurement of the blood oxygenation level‐dependent (BOLD) signal has potential for treatment of neurological disorders and behavioral enhancement. Commonly used methods are based on functional magnetic resonance imaging (fMRI) sequences that sacrifice speed and accuracy for whole‐brain coverage, which is unnecessary in most applications. We present multivoxel functional spectroscopy (MVFS): a system for computing the BOLD signal from multiple volumes of interest (VOI) in real‐time that improves speed and accuracy of neurofeedback. MVFS consists of a FS pulse sequence, a BOLD reconstruction component, a neural activation estimator, and a stimulus system. The FS pulse sequence is a single‐voxel, magnetic resonance spectroscopy sequence without water suppression that has been extended to allow acquisition of a different VOI at each repetition and real‐time subject head motion compensation. The BOLD reconstruction component determines the T2* decay rate, which is directly related to BOLD signal strength. The neural activation estimator discounts nuisance signals and scales the activation relative to the amount of ROI noise. Finally, the neurofeedback system presents neural activation‐dependent stimuli to experimental subjects with an overall delay of less than 1 s. Here, we present the MVFS system, validation of certain components, examples of its usage in a practical application, and a direct comparison of FS and echo‐planar imaging BOLD measurements. We conclude that in the context of realtime BOLD imaging, MVFS can provide superior accuracy and temporal resolution compared with standard fMRI methods. © 2014 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 24, 138–148, 2014     

Neural systems subserving valence and arousal during the experience of induced emotions.

The circumplex model of affect construes all emotions as linear combinations of 2 independent neurophysiological dimensions, valence and arousal. We used functional magnetic resonance imaging to identify the neural networks subserving valence and arousal, and we assessed, in 10 participants, the associations of the BOLD (blood oxygen level-dependent) response, an indirect index of neural activity, with ratings of valence and arousal during the emotional experiences induced by the presentation of evocative sentences. Unpleasant emotional experience was associated with increased BOLD signal intensities in the supplementary motor, anterior midcingulate, right dorsolateral prefrontal, occipito-temporal, inferior parietal, and cerebellar cortices. Highly arousing emotions were associated with increased BOLD signal intensities in the left thalamus, globus pallidus, caudate, parahippocampal gyrus, amygdala, premotor cortex, and cerebellar vermis. Separate analyses using a finite impulse response model confirmed these results and revealed that pleasant emotions engaged an additional network that included the midbrain, ventral striatum, and caudate nucleus, all portions of a reward circuit. These findings suggest the existence of distinct networks subserving the valence and arousal dimensions of emotions, with midline and medial temporal lobe structures mediating arousal and dorsal cortical areas and mesolimbic pathways mediating valence. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

基于小波变换的功能磁共振图像时间序列分步去噪

功能磁共振图像（fMRI）数据中反映大脑神经活动的感兴趣信号常受到结构噪声和随机噪声的影响。为消除上述噪声对分析激活体素的影响,对经过SPM标准预处理的体素时间序列进行Activelets小波变换,并在得到尺度系数及细节系数后,针对两类噪声的不同特点进行分步去噪。第一步,在受结构噪声影响的尺度系数上,选用独立成分分（ICA）析去识别并消除结构噪声源;第二步,提出一种改进的空域相关去噪算法在细节系数上对信号进行处理。值得注意的是,该算法利用邻域体素之间的相似性,判定所处位置的细节系数反映噪声还是神经活动。实验结果表明,经过这两步处理的数据可有效消除噪声的影响,其中框架位移减少了1.5mm,尖峰百分比减少了2%,此外由去噪后的信号获得的脑激活图中一些明显的伪激活区得到抑制。     

基于子图相似性的多动症患者脑网络分析

多动症会严重影响儿童发育，对多动症患者的有效诊断受到广泛关注。该文结合脑网络的拓扑结构信息和图上的信号，提出一种基于稀疏表示的图相似性计算方法，从微观到宏观分析脑区之间的差异。该方法使用Pearson相关系数构建全连通脑网络，基于稀疏表示从底层结构中提取节点子网络，根据图核函数计算子网络相似性，最后给出了脑网络相似性的全局指标。以受试者间的相似性作为特征在公共数据集ADHD-200上的分类实验结果表明，该方法能够以93.1%的准确度区分多动症患者和健康对照者，分类性能明显优于其他已有算法。此外，结果表明多动症患者在中央前回、丘脑、海马和脑岛等脑区之间有更强的连接。