摘要
为了解决脑机接口(BCI)中不同意识任务下脑电信号分类问题,针对运动想象脑电(EEG)的事件相关去同步/同步(ERD/ERS)现象,提出一种基于支持向量机(SVM)的实用分类算法。该算法首先对脑电信号进行滤波,获得对运动想象比较敏感的频段,对滤波后的脑电信号,通过去均值减小由于均值不同所造成的误差,然后,再提取基于ERD/ERS的脑电能量场强特征,对提取的特征,运用支持向量机(SVM)进行分类,得到了满意的效果。结果表明,此方法可为脑机接口技术的应用提供有效的手段。
In order to label the electroencephalogram (EEG) under different imagery task in brain- computer inerface(BCI) technology, a practical classification method was put forward which was based on theory of event - related EEG desychronization/synchronization and support - vector machine (SVM) algorithm. In the algorithm, firstly a filter was used to filter the EEG to obtain the frequency band which was sensitive to imagery task, then the mean value was removed from the EEG to minimize the incidental error caused by it. Finally the field strength characteristics related to ERD/ERS were extracted as a vector and the vector was used by SVM to recognize the pattern of the EEG. The results show that this approach provides an effective way for the application of the brain - computer interface,
出处
《生物医学工程研究》
2007年第1期20-23,32,共5页
Journal Of Biomedical Engineering Research
基金
山东省自然科学基金资助项目(Y2005C68)
关键词
运动想象
脑机接口
事件相关同步/去同步
能量场强
支持向量机
Motor image(MI)
Brain - computer interface (BCI)
Event - related synchronization desychronization(ERS/ERD)
Field strength
Support vector machine (SVM)