基于HMM的低信噪比离子通道信号的恢复及参数估计

细胞膜离子单通道信号是皮安级的随机电流 ,膜片钳技术可以记录这些信号。一般认为它是一阶的、状态有限的 Markov过程。某些种类的离子通道 ,电流信号特别微弱 ,完全淹没在背景噪声中 ,传统的膜片钳技术很难检测到 ,只能运用数学方法恢复和估计。在低采样频率情况下 ,由于混叠效应 ,可认为背景噪声是白色的 ;在高采样频率条件下 (高于奈奎斯特频率 ) ,背景噪声是有色的。本文分别综述了白色背景噪声条件下基于隐式 Markov模型和有色背景噪声条件下基于隐式矢量 Markov模型的低信噪比离子单通道信号的恢复和参数估计 ,主要包括前后向算法、EM算法等

The restoring of low signal noise rate single ion channel signal and its estimation of parameters based on HMM

The signal of single ion channel is stochastic ionic currents on the order of 1pA,which can be recorded by means of patch clamp technique.It is assumed that the signal is generated by a first order,finite state Markov process.Many kinds of ion channels currents are especially more weak and dominated by background noise so that their characteristics cannot be measured with any certainty by traditional patch clamp technique.We now deal with this problem in the way of mathematical analysis.With low sampling frequency (below the Nyquist frequency) the background noise may be considered to be white because of aliasing,while it is colored with high sampling frequency(above the Nyquist frequency).Here we summary the restoring of single channel ionic signal with low signal noise rate (SNR) and its estimation of parameters,respectively based on hidden Markov models under the condition of white background noise and hidden vector Markov models under the condition of colored background noise,mainly including of forward backward procedure and EM algorithm.