基于VR-fNIRS的平静、沮丧和恐惧情绪的分类研究  

Research on VR-fNIRS based classification of calm,frustration and fear

作　　者：许博俊 李梦琪 XU Bojun;LI Mengqi(School of Information Engineering and Automation,Kunming University of Technology,Kunming 650500,China)

机构地区：[1]昆明理工大学信息工程与自动化学院,云南昆明650500

出　　处：《现代电子技术》2023年第22期121-125,共5页Modern Electronics Technique

摘　　要：目前有一些心理情绪,比如沮丧和恐惧,从生理角度可能难以区分,因为它们在唤醒度和效价中十分接近。招募18名男性进行测试研究,让测试者戴上功能性近红外光谱技术(fNIRS)帽,在三个VR场景中交互从而完成任务;然后分析被试对平静、沮丧和恐惧的情绪诱发任务的血氧浓度蛋白(HbO)信号。选取均值、方差、最大值、最小值、均方根常用fNIRS特征,以及近似熵(ApEn)、样本熵(SaEn)和排列熵(PeEn)三类熵特征,通过支持向量机(SVM)、K最近邻(KNN)和决策树(DT)实现了对这三种情绪的分类,并对空间分布和脑地形图也进行了研究。最后,使用均方根特征和SVM分类器获得了最佳的三类分类精度。Some psychological emotions,such as depression and fear,may be difficult to distinguish from each other from a physiological perspective,as they are very close in arousal and potency.18 males are recruited for testing and research.Participants wear functional near-infrared spectroscopy(fNIRS)hats and interact in three VR scenarios to complete the task.Then,the blood oxygen concentration protein(HbO)signals of subjects to calm,depression and fear emotion induced tasks.The mean,variance,maximum,minimum,and root mean square fNIRS features and approximate entropy(ApEn),sample entropy(SaEn)and permutation entropy(PeEn)features are selected.The support vector machine(SVM),K-nearest neighbor(KNN)and decision tree(DT)are used to classify these three emotions,and the spatial distribution and brain topographic map are also studied.The root-mean-square feature and SVM classifier are used to obtain the best classification accuracy of three classes.

关 键 词：虚拟现实 功能性近红外光谱技术 情绪诱发 近似熵 支持向量机 K近邻算法 决策树 脑电图 

分 类 号：TN219-34[电子电信—物理电子学] TP391[自动化与计算机技术—计算机应用技术]