基于音乐诱发脑电的抑郁症识别研究  

Research on Recognition of Depression Based on Music-induced EEG

作　　者：陈亮 郭滨[1] 李沐芳 李哲 CHEN Liang;GUO Bin;LI Mu-fang;LI Zhe(School of Electronics and Information Engineering,Changchun University of Science and Technology,Changchun 130022)

机构地区：[1]长春理工大学电子信息工程学院,长春130022

出　　处：《长春理工大学学报（自然科学版）》2021年第4期77-83,共7页Journal of Changchun University of Science and Technology(Natural Science Edition)

基　　金：吉林省科技厅项目(20200404216YY)。

摘　　要：抑郁症是最常见的精神类疾病之一,临床诊断存在困难,有必要寻找一种客观、高效的方式来辅助抑郁症的快速识别。通过融合中性、负性、正性音乐刺激下的不同脑电图(EEG)数据,提出一种新的抑郁识别方法来区分轻度抑郁症患者和正常对照组。在接受不同音乐刺激的同时同步记录抑郁症患者和正常对照组的脑电信号;然后从各模态的脑电图信号中提取线性和非线性特征,得到各模态的特征;此外,采用线性组合技术融合不同模型的脑电特征,构建全局特征向量,找出最佳的特征子集。最后比较了各分类器K-NN、DT和SVM的分类精度。实验结果表明,基于音乐刺激诱发脑电建立有效的抑郁症识别模型,KNN分类器的分类准确率最高达86.93%,可为抑郁症的辅助识别提供客观的指标和依据。Depression is one of the most common psychiatric diseases.Clinical diagnosis is difficult.It is necessary to find an objective and efficient way to assist the rapid identification of depression.By fusing different electroencephalogram(EEG)data under the stimulation of neutral,negative,and positive music,a new method of depression recognition was proposed to distinguish patients with mild depression from normal control groups.Simultaneously the EEG signals of de-pression patients and the normal control group were recorded while receiving different music stimuli;then the linear and non-linear features were extracted from the EEG signals of each modal to obtain the characteristics of each modal.In addi-tion,linear combination technology was used to fuse the EEG features of different models;a global feature vector was con-structed and the best feature subset was found.Finally,the classification accuracy of each classifier K-NN,DT and SVM was compared.The experimental results showed that an effective depression recognition model was established based on the EEG induced by music stimulation.The classification accuracy of the KNN classifier was up to 86.93%,which can provide objective indicators and basis for the auxiliary recognition of depression.

关 键 词：抑郁症 音乐 脑电信号 分类 

分 类 号：TN911[电子电信—通信与信息系统]