摘要
针对目前脑电功谱计算中因方法不同而造成的彼此间结果难以比较的状况，本文提出了脑电比功谱的概念和计算方法。脑电比功谱即脑电信号的功率在单位频率（每赫）上的分布。计算方法包括如下步骤：（１）脑电幅值的规范化处理；（２）用快速哈特莱或富氏变换求出与所选样本长度有关的各频率节点上的功率值；（３）通过积分处理求出单位频率上的功率；（４）进行对数变换将功率值以ｄＢ表示；（５）校正平均功率水平。对模拟白噪信号及两类典型脑电信号的比功谱计算结果表明，本文提出的方法切实可行，为不同条件下的脑电功谱的比较提供了一个可靠的手段。
In order to solve the problem in attempting to compare the EEG spec-trum results
derived from different laboratories with different recording parameters and EEG sample
length, a concept of EEG specific power(EEG-SP) which is defined as the power per unit
frequency, and the method of its calculation are presented in this paper.The calculating
procedures of EEG-SP are as follows.(1) Reducing the amplitude of recorded EEG signal to the
level at recording electrode so that the effects of gain differ-ence during amplification and A/D
conversion are eliminated.(2) The powers at eachfrequency node for the selected EEG sample
length(e.g., 2 or 4 seconds) are derived us-ing a Fast Hartley Transform algorithm (FHT) which
is equivalent to FFT but more con-venient for real number series.(3) Specific powers are
derived by the proposed integra-tion method.(4) The calculated specific powers are expressed
in dB via a logarithmictransform.(5) The correcting value for mean power is added to
compensate the energyleakage due to limited freguency resolution during FHT calculation. For
convenience ofcalcuiation, sampling rate of 256 or 512 per second during A/D conversion is
recom-mended.As examples ,the specific power spectra(SPS) were calculated by using the
methoddescribed for two types of EEG data with alpha or beta dominance. In each case as
shownin Fig. 3 and Fig.4, the EEG-SPSs calculated with 1,2 or 4 second sample length
arehighly comparable both in the peak positions and the absolute power levels. Therefore,itis
concluded that the method described in the present paper provides a sound basis for
thecomparisons of EEG power spectrum data collected from different laboratories under
dif-ferent conditions and at different times.
出处
《航天医学与医学工程》
CAS
CSCD
1994年第3期169-174,共6页
Space Medicine & Medical Engineering
关键词
脑电图
频谱分析
脑电比功谱
electroencephalogram, power spectrum,
spectrum analysis, EEG spe-cific power spectrum
