摘要
针对脑机接口(brain-computer interface,BCI)系统特征提取较慢的现状,提出基于约束独立分量分析(constrained independent component analysis,cICA)的P300特征提取方法.首先,针对各位P300实验被试,通过EEG图像研究其特有P300时域特性;然后,根据P300特性构建参考信号,并将参考信号与独立分量分析(independent component analysis,ICA)方法结合,基于64导联EEG,提取出与P300相关度最大的独立分量;最后,依据提取出的独立分量构造3维特征向量进行分类.实验采用线性分类器,针对BCI Competition II dataset IIb和BCI Competition III dataset II两组公共数据集进行了验证.结果表明,提出方法在3次叠加平均下识别正确率达67.1%,15次达95.2%,在相同实验条件下,分类时间也较其他方法缩短.
Considering the current time-consuming feature extraction of the brain-computer interface, a feature extraction method based on constrained ICA was proposed for P300-BCI. The temporal P300 character of every subject was studied using the EEG image, and then, reference signals were built according to the temporal P300 character. Using the reference signals combined with ICA, the most correlative independent components were extracted based on 64-channel EEG. According to the extracted independent components, 3-dimensional feature vectors were built and put into the linear classifier at last. Two public datasets of BCI Competition Ⅱ and Ⅲ were used to verify the method. The results show that the recognition accuracy can be improved to 67.1% only with three times average, and to 95.2% with fifteen times average. The computation time is also shorter than other methods in the same experimental conditions.
出处
《东北大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2014年第3期419-422,437,共5页
Journal of Northeastern University(Natural Science)
基金
国家自然科学基金资助项目(61071057)
关键词
脑机接口
脑电
特征提取
约束独立分量分析
识别正确率
brain-computer interface
electroencephalogram （ EEG ）
feature extraction
constrained ICA
recognition accuracy
