摘要
目的:为了获得癫痫脑电的某些非线性特征,研究了癫痫患者脑电的信息熵,并和健康人脑电的信息熵作比较。方法:采用基于相空间重构脑电信号的新方法来计算脑电信息熵。分别计算了100例癫痫患者和50例健康者的脑电样本。结果:癫痫脑电的平均信息熵为1.696。而健康者脑电的平均信息熵为1.920。值得注意的是,随着嵌入维数的增加,癫痫脑电的信息熵呈现出逐渐减小的趋势,而健康脑电却正好相反,呈现出逐渐增大的趋势。结论:癫痫脑电在低维空间的信息熵比高维空间的大,而健康脑电的情况正好相反。
AIM:To study the information entro py of e pileptic electroencephalogram(EEG)and compare with entropy of normal EE G for obtaining some non-linear features of epileptic EEG.METHODS:One hundred samples of epileptic EEG and 50samples of normal EEG were calculated respecti vely by using a novel method that was based on the reconstruction of EEG in the phase space.RESULTS:The results showed that the mean valu e of information en tropy was 1.696in epileptic EEG ,and 1.920in normal EEG.The notable thing was an increasing in embedding dimen sion results in a decreasing in entro py of epileptic EEG but oppositely for n ormal EEG.CONCLUSION:The information entropy of epileptic EEG is larger when in the low-dimension space than in the high-dimension space,and oppositely for the normal EEG.
出处
《中国临床康复》
CSCD
2003年第24期3304-3305,共2页
Chinese Journal of Clinical Rehabilitation
