手指活动相关脑磁图的时-频域特征分析

Finger movement related magnetoencephalography signal processing based on time frequency representation of power

对皮层神经电生理信号进行时频分析是准确描述其信号特征的方法。为了研究手指活动相关脑磁图的时-频域分布特征,通过计算多通道脑磁图数据的时频功率表示(time frequency representations of power,TFRs),得到各通道数据在时频面上的能量分布;同时设计了一个基于Visual Basic和MATLAB的离线脑磁图数据处理的可视化软件以规范时频特征分析处理的流程。对一癫痫患者右手食指运动时记录的脑磁图数据进行处理,结果显示,患者对侧中央区和顶区皮层被显著激活,同侧中央区也有少数通道被激活。并且发现,皮层活动的10~25 Hz频率成分首先在–0.5^+0.3 s时间段表现为功率抑制,接着有强烈的功率增强活动显示,最显著的功率增强集中在0.4~0.7 s区间。初步的实验结果得到了手指活动相关神经活动在时-频域的表现,同时也为研究皮层活动在时域、频域以及空间的分布特征提供了方法。

Time-frequency processing of brain electrical and magnetic signals can reveal a more detailed picture of cortical activities. To investigate brain activities associated with self-paced finger tapping task in temporal-frequency domain, time frequency representations of power （TFRs） are calculated for all the MEG channels and the results are presented as the powers at each frequency bin and each time bin of interest for each channel in the present work. A visual software based on Visual Basic and Fieldtrip MATLAB toolbox was designed to process the MEG signal offline. A data set recorded in a trial test from an epileptic patient performing a self-paced tapping with right index finger was analyzed using this software. Contralateral central and parietal parts of the cortex as well as several ipsilateral channels in central area are remarkably activated. Brain activity components between 10 to 25 Hz are firstly suppressed in -0.5 to ＋0.3 second, and afterwards burst activity is displayed with a weaken period around 1 second. Most significant activation could be observed from 0.4 to 0.7 second. More internal relations in time and frequency domains of cortical neural activity in motor control could be recognized from analyzing TFRs of more normal subjects＇ experimental data sets.