The phase lock value(PLV) is an effective method to analyze the phase synchronization of the brain, which can effectively separate the phase components of the electroencephalogram (EEG) signal and reflect the influence of the signal intensity on the functional connectivity. However, the traditional locking algorithm only analyzes the phase component of the signal, and can't effectively analyze characteristics of EEG signal. In order to solve this problem, a new algorithm named amplitude locking value (ALV) is proposed. Firstly, the improved algorithm obtained intrinsic mode function using the empirical mode decomposition, which was used as input for Hilbert transformation (HT). Then the instantaneous amplitude was calculated and finally the ALV was calculated. On the basis of ALV, the instantaneous amplitude of EEG signal can be measured between electrodes. The data of 14 subjects under different cognitive tasks were collected and analyzed for the coherence of the brain regions during the arithmetic by the improved method. The results showed that there was a negative correlation between the coherence and cognitive activity, and the central and parietal areas were most sensitive. The quantitative analysis by the ALV method could reflect the real biological information. Correlation analysis based on the ALV provides a new method and idea for the research of synchronism, which offer a foundation for further exploring the brain mode of thinking.
锁相位算法是分析脑区同步特性的有效方法。基于传统锁相位算法只针对脑电信号的相位分量进行分析，而不能有效地分析脑电信号的其他特征分量的问题，本文提出了一种改进锁相位算法。该算法首先基于经验模态分解获得固有模态函数，以此作为希尔伯特变换的输入求出所需瞬时幅值，计算锁幅值。基于此算法，不仅可以度量脑电信号采集位点之间的变化情况，而且可以度量各个位点自身的脑电信号瞬时振幅的变化情况，从而达到提取脑电信号同步特征的效果。本文采集了 14 名受试者在不同难度认知任务下的脑电信号，并基于改进锁相位算法，分析大脑在完成认知任务过程中各脑区之间的同步性。结果表明，大脑幅值同步程度与认知活动之间呈负相关，且大脑中央区和顶叶最为敏感。基于改进锁相位算法进行的同步性量化，能够真实地反映数据之间的生物信息，为大脑同步性研究提供了一种新的方法，为更好地探索脑区的相关性与同步性提供新思路。.
Keywords: amplitude locking value; brain function correlation; electroencephalogram signal; empirical mode decomposition; phase locking value.