基于机器学习的抑郁症脑电识别研究  

Machine learning-based EEG recognition study for depression

作　　者：骆睿鹏 邹任玲[1] 孟令鹏 谈宏伟 刘巨涛 徐澜菲 胡秀枋[1] 曹立 LUO Ruipeng;ZOU Renling;MENG Lingpeng;TAN Hongwei;LIU Jutao;XU Lanfei;HU Xiufang;CAO Li(School of Health Science and Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China;Department of Neurology,Shanghai Sixth People′s Hospital,Shanghai Jiao Tong University School of Medicine,Shanghai 200233,China)

机构地区：[1]上海理工大学健康科学与工程学院,上海200093 [2]上海交通大学医学院附属第六人民医院神经内科,上海200233

出　　处：《智能计算机与应用》2023年第11期82-87,共6页Intelligent Computer and Applications

基　　金：上海市科技创新行动计划产学研医合作领域项目(21S31906000)。

摘　　要：抑郁症已成为全球主要的健康负担,有效检测抑郁症是公共卫生的一大挑战。为了更好地识别抑郁症,本文提取了不同类型的脑电信号特征,其中包括线性特征和非线性特征,对抑郁症患者的脑电信号进行综合分析。并使用不同的机器学习算法模型分类器来评估最优特征集。在结合所有类型的特征对MDD患者进行分类时,获得了最佳识别准确率达到91%左右。这项基于机器学习和脑电信号(EEG)的抑郁症识别研究为未来应用于抑郁症领域的早期筛查、辅助诊断以及辅助治疗决策等方面提供了一种辅助方案和一定的参考价值。Depression has become a major global health burden,and effective detection of depression is a major public health challenge.To better identify depression,the paper extracts different types of EEG signal features,which include linear features and nonlinear features,for the comprehensive analysis of EEG signals from depressed patients.And different machine learning algorithms are used to model classifiers to evaluate the optimal feature set.The best identification accuracy of about 91%is obtained when combining all types of features to classify MDD patients.This study on the identification of depression based on machine learning and EEG signals provides a complementary solution and the reference value for future applications in the field of depression for early screening,aiding diagnosis and assisting treatment decisions.

关 键 词：脑电信号 抑郁症 静息态 刺激态 特征提取 机器识别 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]