摘要
连续循环平均反卷积(Continuous loop averaging deconvolution,CLAD)是近年来用于提取高刺激率模式下听觉诱发电位(Auditory evoked potential,AEP)的一种行之有效的方法。但是,CLAD方法在频率域求解时,对刺激序列的频谱特性有严格的限制,给应用带来不便和局限。本文提出一种在时域实现反卷积的方法,将其转化为线性变换矩阵的逆滤波处理。并且利用奇异值分解分析了由不良序列带来的不适定问题,引入正则化技术改善病态矩阵对重建结果的影响。最后比较了若干种典型刺激序列和不同噪声条件下AEP的恢复实验,结果表明本方法可以较好地解决不良序列和一般噪声水平条件下暂态AEP信号的恢复重建。
Continuous loop averaging deconvolution（CLAD）is a recently developed method to restore the auditory evoked potential（AEP）under high stimulus rate condition.This method solves the deconvolution problem in frequency domain for computational efficiency,but suffers from stringent limitation in selecting a stimulus sequence with required spectral property.Hereby we propose a new method to solve the deconvolution problem in time domain by constructing a linear transform matrix to model the convolution process.To understand the AEP distortion caused by the ill-posed matrix generated from a bed stimulus sequence,we assess the matrix property using singular value decomposition（SVD）technique and introduce Tikhonov regularization method to deal with the ill-posedness.In the stimulation experiment,we compare some typical sequences with different ill-posedness conditions and restore the transient AEPs under various noise levels.These results justify the proposed approach to the AEP deconvolution with less restriction on the sequence selection.
出处
《数据采集与处理》
CSCD
北大核心
2015年第5期1011-1019,共9页
Journal of Data Acquisition and Processing
基金
国家自然科学基金(61172033
61271154)资助项目
关键词
听觉诱发电位
高刺激率
反卷积
正则化技术
逆滤波
auditory evoked potential
high stimulus rate
deconvolution
regularization techniques
in-verse filtering
