基于自适应L滤波器的脑事件关联电位单次提取

Estimation of Single Event Related Potentials Using Adaptive L-Filtering

提出了一种基于自适应L滤波的脑事件关联电位单次提取方法 .将背景噪声看成一系列起伏状脉冲 ,利用零交叉方法估计各单次记录信号中噪声的脉冲宽度 ,取其最大值作为基准参数 ,自适应地动态确定各单次滤波时所用窗口的大小 .在窗口大小确定后 ,对各单次记录到的信号进行L滤波处理 .实验所采用的刺激方式为视频奥德勃 (odd ball)模式 ,直接对真实数据进行处理 .将实验结果与普通叠加平均结果及原始单次记录进行了对此 ,结果表明脑事件关联电位单次提取效果良好 ,所提出的方法对噪声的分布和平稳性没有特殊要求 ,因而具有较强的适应性和稳健性 .

A new acquiring method based on adaptive L-filtering (ALF) that is used for estimation of single event related potentials (sERP) is presented. The background noise is considered as a series of impulse and analyzed by zero-crossing method. The length of L-filter is adaptively set according to the maximum width of the impulse noise for each recorded raw data. The data for experiments were measured with visual odd-ball trial. The experimental results show that the performance of sERP estimation is excellent and the proposed method is robust enough to deal with sERP processing.