摘要
应用Ｇｒａｓｓｂｅｒｇｅｒ—Ｐｒｏｃａｃｉａａｌｇｏｒｉｔｈｍ算法，通过点列长度、时间延迟、嵌入维数等参数的变化，对人的８导脑电时间序列的相关维数值的影响进行了研究。结果表明不仅参数的选用是否合理对计算值影响很大，而且脑电信号不像是一个稳定的混沌吸引子，因此拟用相关维数作为一项客观指标来准确地表示大脑的不同功能状态，几乎是不可能的。
Grassberger Procaccia algorithm is used to calculate the correlation dimension of 8-channel human EEG signals. T'he influences of length of time series, time delay and embedding dimesion on the value of EEG correlation dimension were investigated. It indicated that the results of calculation are greatly influenced by the reasonable choice of parameters. Besides, it seemed that EEG signals can not be defined as a stable chaotic attractor. Therefore,the attempt to take the correlation dimension as a precise and objective parameter to indicate human brain functions in different physiological states is almost unfeasible.
出处
《生物物理学报》
CAS
CSCD
北大核心
1995年第1期49-52,共4页
Acta Biophysica Sinica
关键词
脑电
相关维数
时间延迟
嵌入维数
脑电图
Electroencephalogram
Correlation dimension
Time delay
Embedding dimension
