摘要
为了克服反混叠滤波器和有色背景噪声的影响 ,本文提出了一种自适应算法 ,估计离子通道的动力学特征参数和背景噪声的统计特征 ,然后基于这些参数 ,运用统计技术从强噪声的膜片钳记录中恢复离子通道信号。这种算法交叉耦合了递归的 expectation- m axim ization(EM)算法和递归扩展最小二乘算法。递归 EM算法最优地估计隐Markov模型参数 ,递归扩展最小二乘算法最优地估计背景噪声的特征。仿真研究表明这种交叉耦合算法一致收敛 。
In order to overcome the effects of the anti-aliasing filter and the colored background noise, an adaptive algorithm is proposed to estimate the parameters of ion channel kinetics and the background noise, and whereafter the ion channel signal could be restored from the strong noisy patch-clamp recordings. The algorithm cross-couples the recursive expectation-maximization algorithm, which estimates optimally the parameters of hidden Markov model, and the recursive extended least square algorithm, which estimates optimally the characteristics of the background noise. Simulation suggests that this cross-coupling algorithm convergences consistently, and is very robust to the inexact conformation number.
出处
《生物医学工程学杂志》
EI
CAS
CSCD
2002年第3期444-448,共5页
Journal of Biomedical Engineering