基于统计分布熵的抑郁症脑电信号分析

Analysis of Depression Electroencephalogram Based on Statistics Distribution Entropy

针对目前抑郁症研究和诊断中量化分析参数和分析方法不足的情况,本文提出和定义一种能对脑电活动的状态分布进行描述、并能用来计算和区分抑郁症患者和正常人脑电活动差异的统计分布熵方法。应用该方法对抑郁症患者和正常对照组的脑电信号统计分布熵进行数值计算,然后分析讨论它们之间的差异,最后对结果进行了统计分析。实验结果表明:抑郁症患者脑电的状态分布熵在部分脑区显著高于正常健康人,表现出较强的差异性。证明该统计分布熵能够表征大脑电活动的分布状态,提供反映其活动是否发生异变的信息,可以作为度量大脑电活动分布状态和分析脑电信号是否异常的一个物理参数。这对其用作诊断其他脑精神疾病的物理指标也具有积极意义。

A method is proposed to calculate and analyze electroencephalogram to improve the situation in which there is an emergency for the effective quantitative parameters for mental disorders. The method first defines a statistics distribution entropy to describe the state distribution of brain electrical activity, which can calculate and analyze the state difference of it. The entropy is applied to numerical calculation of electroencephalogram signal between depression patients and normal control group. Meanwhile, the difference is compared between them. The experiment shows that the statistics distribution entropy in depression patients is significantly greater than that of the normal healthy people in some brain regions. Further analysis proves that the entropy can be used as a parameter to measure the state distribution of brain electrical activity and to analyze its difference. The analysis also tells that the entropy plays an important role in diagnosis of other mental disorder.