消除脑电图伪差的数据处理方法

A data processing method for eliminating EEG artifacts

本文提出了一种消除脑电图主要伪差（肌电伪差、眼动伪差和基线漂移）的实用有效方法，即用线性Ｋａｌｍａｎ滤波来消除脑电图的肌电伪差，用ＦＩＲ数字高通滤波器抑制眼动伪差及基线漂移，取得了良好的效果，为进一步设计脑电图的计算机自动分析和诊断系统，解决了关键性的问题。在数据处理的同时还对线性Ｋａｌｍａｎ滤波进行了改进，即去除了肌电伪差模型，使系统更为简洁，滤波效果更好。

This paper describs an effaective method for removing main artifacts (the muscle artifacts, the ocular artifacts and the DC drift) from the EEG signal. In procedure, the muscle artiffacts is reduced by the linear Kalman filter, the eye movement artifacts & the DC drift are reducted by the high-pass FIR digital filter. The key of designing the automaticanalysis and diagnosing system for EEG is solved. Meanwhile, the author has improved the method for removing the muscle artifacts-Kalman filter, : the EMG artifacts model is eliminated. Thus, the system is more simple and useful and may find general application in the field of noise reduction contained in the wide set of biological signals