EEG信号结合特征融合技术诊断精神分裂症和抑郁症  

EEG signals combined with feature fusion in the diagnoses of schizophrenia and depression

作　　者：吴恒[1] 刘浩 肖萌[3] 肖开提·苏理旦[3] WU Heng;LIU Hao;XIAO Meng(Affliated Cancer Hospital of Xinjiang Medical University,Urumqi 830000,China)

机构地区：[1]新疆医科大学附属肿瘤医院,新疆乌鲁木齐830000 [2]新疆大学软件学院 [3]乌鲁木齐市第四人民医院

出　　处：《精神医学杂志》2024年第2期176-180,共5页Journal of Psychiatry

基　　金：乌鲁木齐市卫生健康委科技计划项目(编号:202030)。

摘　　要：目的探索通过机器学习算法结合脑电信号实现对精神分裂症和抑郁症的诊断。方法分别采集33例精神分裂症患者和28例抑郁症患者的脑电信号,并将采集到的脑电图信号格式由EDF格式转化为ASCII格式,提取脑电信号的Lempel-Ziv复杂度、最大李雅普诺夫指数、Higuchi分形维数等特征。应用特征融合策略对特征进行融合,形成新的特征向量,然后利用机器学习分类算法进行分类研究。结果最终基于高斯核函数的支持向量机(SVM)的分类准确率为84.85%,其中灵敏度为89.47%,特异性为78.57%。结论通过提取EEG脑电信号特征结合机器学习算法对精神分裂症和抑郁症进行识别,对开发新型的精神分裂症和抑郁症的诊断技术具有一定的研究意义。Objective To explore machine learning algorithms combined with EEG signals in the diagnoses of schizophrenia and depression.Methods EEG signals from 31 patients with schizophrenia and 28 patients with depression were selected and converted from EDF format to ASCII format.Features such as Lempel-Ziv complexity,maximum Lyapunov exponent,and Higuchi fractal dimension were extracted from the EEG signals.These features were then fused using a feature fusion strategy to create new feature vectors,which were classified through machine learning algorithms.Results The classification accuracy of the support vector machine(SVM)with a Gaussian kernel was 84.85%,with a sensitivity of 89.47%and a specificity of 78.57%.Conclusion This study demonstrates the potential of combining EEG signal feature extraction with machine learning algorithms for the identification of schizophrenia and depression,offering valuable insights for developing novel diagnostic techniques for these mental disorders.

关 键 词：精神分裂症 抑郁症 脑电信号 机器学习 特征融合 

分 类 号：R749.3[医药卫生—神经病学与精神病学] R749.4[医药卫生—临床医学]