基于少量导联的脑-机接口实现方法研究  

The Study on the Method of Brain-computer Interface with Few-channel Electroencephalogram

作　　者：施晨韵 吴小培[1] 殷婵 欧阳蕊 SHI Chenyun;WU Xiaopei;YIN Chan;OUYANG Rui(College of Computer Science and Technology,Anhui University,Hefei 230601,China)

机构地区：[1]安徽大学计算机科学与技术学院,合肥安徽230601

出　　处：《皖西学院学报》2019年第2期36-42,共7页Journal of West Anhui University

基　　金：国家自然科学基金资助(61271352);安徽大学2017年大学生科研训练计划项目(KYXL2017061)

摘　　要：针对少量导联脑电信号在脑-机接口实现中的可靠性和适用性问题,提出了一种基于独立分量分析的运动想象脑-机接口实现方法,系统输入为运动皮层三导联脑电信号。为了保证在少量导联情况下的算法稳定性和可靠性,提出采用自测试方法对空域滤波器进行优化设计。基于脑-机接口国际竞赛数据集对所提方法进行验证,所得运动想象分类结果优于竞赛优胜者的分类结果。Aiming at the realizability and applicability of brain-computer interface(BCI)system with few-channel electroencephalogram(EEG),an implementation method of motor imagery BCI system based on independent component analysis(ICA)was proposed in this paper.Three channel EEG signals from the motor cortex were employed as input of the MI-BCI system.In order to ensure the stability and reliability of BCI system,the ICA spatial filter was optimized with self-test technique.The proposed method was validated on the public benchmark dataset(BCI competition IV-2 b).The experiment results indicated that the proposed method outperformed that used by the winner in BCI competition in terms of classification accuracies.

关 键 词：脑-机接口 运动想象脑电 独立分量分析 空域滤波 特征提取 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]