摘要
许多生物电信号的获取过程中都夹杂了脉搏伪迹,表面肌电信号(SEMG)也是一样。本文通过相邻部位同时采集两路SEMG,一路为待处理信号,另一路作为参考信号。采用小波变换提取参考SEMG中的脉搏波,与待处理的SEMG构建独立分量分析的输入,最后用FastICA算法分离出待处理SEMG中的脉搏波,得到去除脉搏伪迹的SEMG。实验结果表明,该方法用于SEMG中的脉搏伪迹的消除是非常有效的。
Many bioelectric signals are often mixed with pulse artifacts while being acquired, the same as Surface Electromyography （SEMG）. In this paper, one pair of SEMG signals are recorded from an adjacent part, one for objective signal, another for referenced signal. By using wavelet transform, the pulsewave is extracted first from the referenced SEMG signal, which forms the input of independent component analysis （ICA） together with objective SEMG signal. Finally, the pulse-wave is removed from objective SEMG signal by FastICA algorithm, and the SEMG signal without pulse artifacts is obtained. The experimental results indicate that this method is an effective way to remove pulse artifacts from SEMG signals.
出处
《传感技术学报》
CAS
CSCD
北大核心
2008年第9期1498-1502,共5页
Chinese Journal of Sensors and Actuators
基金
863项目资助(2008AA04Z212)
浙江省科技计划项目资助(2007C23088)
