【摘要】 临床上分析癫痫脑电信号非常重要。由于临床记录的癫痫脑电信号中含有大量的伪迹干扰,特别是肌电伪迹,所采集的脑电信号无法正确反映大脑的生理及病理状况。本研究利用小波变换的多分辨率特性和独立分量分析(ICA)的盲源分离特性,把用连续小波变换分解的脑电子带信号作为ICA输入,经ICA分离后,有效地消除了癫痫脑电中的肌电伪迹,并分离出了癫痫样特征波,效果理想。更多还原

【Abstract】 Analysis of epileptic waves in EEG is particularly helpful in clinical field.There are plenty of artifacts contained in almost epileptic waves in EEG recorded in hospital or lab,especially muscle artifacts,which covered the real uesful pathological informations in EEG about the epilepsy.So,with the wavelet transform′s multi-distinguish ability with the independent component analysis′ blind source separation ability,considering the diverse frequency bands as the input of independent component analysis(ICA),after ICA processing,the muscle artifacts was eliminated from the EEG and the epileptic sample waves was separated successfully.更多还原