脑电活动混沌理论预报癫痫发作的可行性

Chaos analysis of electroencephalogram data for predicting epilepsy

目的:应用脑电波变化的非线性动力学特征,以混沌理论对长时程视频脑电资料进行分析,以期预测癫痫发作。方法:选择2004-08/11经解放军第四军医大学附属唐都医院神经外科确诊的癫痫患者13例,患者知情同意。对患者长时程视频脑电资料中32次自发性发作进行分析,比较发作前期与发作间期的相关维数,差异有显著性意义时,即有明显的发作前期。结果:13例患者全部进入结果分析。应用混沌理论分析结果,32次发作中27次发作前相关维数降低,占84%。在这27次发作中,平均可提前(12.26±2.35)min得到预测。13例患者中有11例出现过明显的发作前期(85%),单个患者可预测的发作比例平均为(79±36.6)%。结论:脑电非线性动力学改变在癫痫发作前期较为普遍,运用混沌理论分析脑电资料可以初步预测癫痫的发作。

AIM:To analyze the long term video electroencephalogram data with chaos theory according to non linear dynamical character of the changes of brain wave so as to predict epileptic seizure. METHODS:A total of 13 patients with seizure were selected from the Department of Neurosurgery,Tangdu Hospital of Fourth Military Medical University of Chinese PLA from August to November 2004.Informed consent was obtained from all the subjects before enrollment.Long term video electroencephalogram data of the 32 time spontaneous seizure in patients were analyzed.Correlation dimension analysis method was used to compare the difference between prophase and interphase of epileptic seizure,and significant difference would indicate obvious prophase of epileptic seizure. RESULTS:All 13 patients were involved in the result analysis. The chaos analysis indicated that correlation dimension decreased in 27 of 32 times of spontaneous epileptic seizure (84% ).The mean anticipation time for the 27 time seizures was(12.26± 2.35) minutes.A significant prophase state was detected in 11 of 13 patients(85% ).The mean percentage of anticipated seizures per patient was(79± 36.6) % . CONCLUSION:The non linear dynamics changes of electroencephalogram in prophase of epileptic seizure are a general phenomenon.Anticipating the epileptic seizure with chaos analysis is feasible.