基于独立分量优化子带特征的三类运动想象分类

发布时间：2018-06-16 05:43

本文选题：脑-机接口 + 独立分量分析　；参考：《生物医学工程学杂志》2016年02期

【摘要】：在基于头皮脑电(EEG)信号的脑-机接口(BCI)研究中,用户个体差异性和背景噪声的复杂性是影响BCI系统稳定性的两个主要因素。因此需要针对不同个体进行BCI系统参数优化,其中包括对时域、空域滤波器参数的优化设计和分类器参数的学习。本文以提高BCI系统的准确性为目标,提出了一种结合独立分量分析空域滤波器(ICA-SF)优化设计和EEG多子带特征的BCI信息处理新方法。基于所提方法,对4位受试者在不同时间采集的三类运动想象EEG(MI-EEG)进行分析。实验结果表明,在同一受试者的自交叉测试和不同受试者数据集之间的互交叉验证中,多子带特征结合方法所得到的平均识别率比仅使用单频带所得的平均识别率普遍提高,识别率最大提升可达6.08%和5.15%。
[Abstract]:In the research of brain-computer interface (BCI) based on scalp EEG signal, the difference of user and the complexity of background noise are two main factors that affect the stability of BCI system. Therefore, it is necessary to optimize the parameters of BCI system for different individuals, including the optimization design of the parameters of time-domain and spatial filters and the learning of classifier parameters. In order to improve the accuracy of BCI system, this paper presents a new BCI information processing method which combines the optimization design of independent component analysis (ICA) spatial domain filter (ICA-SF) and the multi-subband feature of EEG. Based on the proposed method, three kinds of motion imagination (EEGMI-EEGG) collected by four subjects at different time were analyzed. The experimental results show that the average recognition rate obtained by the multi-subband feature combination method is generally higher than the average recognition rate obtained by using only one frequency band in the self-crossover test of the same subject and the cross-validation between different data sets. The recognition rate was increased by 6.08% and 5.15% respectively.
【作者单位】：安徽大学计算智能与信号处理教育部重点实验室;安徽大学信息保障技术协同创新中心;
【基金】：国家自然科学基金资助项目(61271352;61401002)
【分类号】：R338;TN911.7

【相似文献】