基于JSD和LMCD的脑电信号分析

发布时间：2018-05-20 19:04

本文选题：脑电信号 + 统计复杂度　；参考：《南京邮电大学》2014年硕士论文

【摘要】：脑电信号的生理机制分析对于评估大脑机能的活跃程度以及健康状态具有重要的意义。本文提出了两种新的基于统计复杂度的脑电信号分析方法，期望能够对传统临床脑电信号的病理分析起到一定的参考价值。本文的工作主要有以下两点：一、基于LMCD癫痫异常脑电信号统计复杂度分析本文介绍了另一种基于欧几里得空间距离定义的差异，应用基于LMCD差异的统计复杂度算法对脑电信号进行分析研究。对癫痫发作时采集的异常脑电信号及正常脑电信号中提取的β波进行复杂度数值计算，结果是癫痫患者的异常脑电信号中β波的平均统计复杂度显著高于正常人的，证明基于LMCD可以作为衡量脑电信号是否异常的参数。二、基于统计复杂度算法的少中年脑电信号分析及对比首先，，介绍了一种用于度量时间序列概率分布之间的差异（散度）：Jensen-ShannonDivergence（JSD）。本文介绍的JSD算法，在“非平衡项”的概念基础上，从而得出了基于JSD统计复杂度的算法，并利用统计复杂度估计时间序列随机本质，该算法可以证明脑电信号在不同年纪段具有明显差异。最后应用这个对少中年脑电信号进行数值计算及对比，结果是中年人的统计复杂度显著高于年少年，表明基于JSD算法的统计复杂度可以有效地区分这两个年级段的脑电信号，并可以作为评估脑功能活跃程度的重要指标。为了把我们研究的算法实现临床使用，辅助医生诊断癫痫疾病，我们在安卓系统中把上述算法进行了实现。
[Abstract]:The analysis of the physiological mechanism of EEG plays an important role in evaluating the activity of brain function and the state of health. In this paper, two new methods of EEG analysis based on statistical complexity are proposed, which are expected to play a certain reference value in the pathological analysis of traditional clinical EEG signals. The main work of this paper is as follows: 1. Statistical complexity analysis of abnormal EEG signals based on LMCD In this paper, another kind of Euclidean spatial distance definition is introduced. The statistical complexity algorithm based on Euclidean difference is used to analyze the EEG signal. The complexity of abnormal EEG signals collected during seizures and 尾 waves extracted from normal EEG signals were calculated. The results showed that the average statistical complexity of 尾 waves in abnormal EEG signals of epileptic patients was significantly higher than that in normal subjects. It is proved that LMCD can be used as a parameter to measure whether EEG signal is abnormal or not. Second, the analysis and comparison of EEG signals in middle and young age based on statistical complexity algorithm. Firstly, a new method to measure the difference between probability distributions of time series is introduced. The JSD algorithm introduced in this paper is based on the concept of "non-equilibrium term", and an algorithm based on the statistical complexity of JSD is obtained, and the stochastic nature of time series is estimated by statistical complexity. The algorithm can prove that there are obvious differences in EEG signals at different ages. Finally, this method is used to calculate and compare the EEG signals of young and middle-aged people. The results show that the statistical complexity of middle-aged people is significantly higher than that of teenagers, which indicates that the statistical complexity based on JSD algorithm can effectively distinguish the EEG signals between the two grades. It can be used as an important index to evaluate the degree of brain function activity. In order to realize the clinical use of the algorithm we studied to assist doctors in the diagnosis of epilepsy, we implemented the algorithm in Android.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R741.044;TN911.6

【参考文献】