摘要
符号动力学分析是脑电分析的一个新的研究方向,符号熵可以较好地反映非线性信号的复杂程度,具有简单、稳定的特点。本文提出了一种新的符号化方法——差分符号化,即在观测数据的切空间中进行符号动力学分析,并进一步比较了不同符号化参数对符号熵的影响。通过对不同生理状态下的脑电信号数据的对比分析表明,应用此方法可以显著地区分出正常与癫痫及睁眼与闭眼等病理及生理脑电信号的复杂度变化情况,对建立客观的脑电信号评价标准、准确进行定量分析具有重要的意义。
Symbolic dynamics may be a new research direction for electroencephalogram（EEG） signal analysis.Symbolic entropy can reflect the degree of complexity of nonlinear signal simply and reliably.In this paper,we propose a new method to symbolize the EEG signal,namely difference symbolization,by which we can analyze the characteristics of dynamics on the tangent space of observation data.Furthermore,we compare and analyze the value of symbolic entropy by choosing difference symbolization parameter.The comparative analyses on different physiological state of EEG data sets show that this method can clearly distinguish the EEGs＇ complexity between normal and epilepsy,between eyes open and eyes closed,and so on.And it is of significance to the establishment of objective criteria for evaluation and fine quantitative analysis of EEG.
出处
《生物医学工程学杂志》
EI
CAS
CSCD
北大核心
2010年第2期407-410,共4页
Journal of Biomedical Engineering
关键词
脑电图
符号动力学
符号时间序列分析
复杂度
符号熵
Electroencephalogram（EEG）
Symbolic dynamics
Symbolic time series analysis
Complexity
Symbolic entropy
