基于近似熵快速算法的静息态脑磁信号分析

发布时间：2018-05-15 05:07

本文选题：脑磁信号 + 小波变换　；参考：《华中师范大学学报(自然科学版)》2017年03期

【摘要】：为了探究静息态精神分裂症患者脑磁信号的非线性动力学特性,提出了一种将小波变换和近似熵相结合的特征提取方法.该方法首先通过小波变换,将10个正常人和10个精神分裂症患者的脑磁信号进行6层小波分解,提取对应于脑磁信号θ波段和α波段的小波系数,继而计算和比较两类人近似熵的分布情况.实验结果表明,相同情况下精神分裂症患者MEG信号的各脑区和各通道间的近似熵都普遍高于正常人,α波段的额叶和中央区域尤为突出.该结果为进一步研究患者MEG信号特征进而建立相应的分类诊断模型提供了思路.
[Abstract]:In order to explore the nonlinear dynamic characteristics of brain magnetic signals in patients with resting schizophrenia, a feature extraction method combining wavelet transform and approximate entropy is proposed. In this method, the brain magnetic signals of 10 normal persons and 10 schizophrenic patients are decomposed into six layers by wavelet transform, and the wavelet coefficients corresponding to theta and 伪 bands of the brain magnetic signals are extracted. Then the distribution of the approximate entropy of two kinds of people is calculated and compared. The experimental results show that the approximate entropy of MEG signals in schizophrenia patients is generally higher than that in normal controls under the same conditions, especially in the frontal and central regions of 伪 band. The results provide a way to further study the characteristics of MEG signal and establish the corresponding classification diagnosis model.
【作者单位】：上海海事大学信息工程学院;
【基金】：第48批教育部留学回国人员科研启动基金项目
【分类号】：R749.3

【相似文献】