诱发电位快速提取算法的新进展

诱导电位（ＥＰ）和事件相关电位（ＥＲＰ）的单次提取是生物医学信号处理领域颇受关注的一个研究问题。本文综述了近年来的ＥＰ／ＥＲＰ快速提取余步垢一些新进展，主要介绍了子窨正则化、自适应期斯径向基神经网络、独立分量分析和划性检测四种方法，这些方法都立足于用较少的刺激次数提取出较高质量的诱发电位或事件相关电位，展示了ＥＰ／ＥＲＰ快速提取的前景。     

基于参数模型和独立分量分析的事件相关诱发电位单次提取

我们针对脑电事件相关电位（ERP）这种信噪比极低的信号检测问题，提出了两种ERP信号单次提取方法，能非常有效地同时去除自发脑电、眼动伪迹和工频噪声三种常见噪声。（1）首次对自发脑电、眼动伪迹和工频噪声这三种常见成分连同事件相关电位同时进行ARX建模，利用基于最小二乘（Ls）的ARX算法进行参数辨识获得提取结果；（2）利用独立分量分析，采用FastICA算法进行事件相关电位的提取。明确指出ICA分解的一些重要分解特性及其内在机理，针对实际情况对FastICA算法进行了改进，实现了分解结果对ERP成分的自适应映射。数值仿真实验结果表明两种方法均有较高的信号分解提取能力。     

诱发电位快速提取算法的新进展

诱发电位 (EP)和事件相关电位 (ERP)的单次提取是生物医学信号处理领域颇受关注的一个研究问题。本文综述了近年来 EP/ERP快速提取算法的一些新进展 ,主要介绍了子空间正则化、自适应斯径向基神经网络、独立分量分析和奇异性检测四种方法 ,这些方法都立足于用较少的刺激次数提取出较高质量的诱发电位或事件相关电位 ,展示了 EP/ERP快速提取的前景。     

基于小波和卡尔曼平滑的事件相关电位单次提取

本研究提出一种事件相关电位单次提取方法,可有效减少实验次数,并可探索实验之间ERP的变异性。此方法基于小波和卡尔曼平滑,首先利用小波变换考察ERP平均信号的时频特性,根据ERP不同分量出现的时间位置,在不同尺度上选取特定的单次实验ERP小波系数构成观测向量,其为真实ERP小波系数状态向量与噪声之和,然后对观测向量进行卡尔曼平滑,最后对卡尔曼平滑后的小波系数进行小波重构,得到单次提取的ERP信号。仿真实验表明,基于小波和卡尔曼平滑的方法不仅信噪比提高约16~18 dB,优于30次叠加平均、简单小波方法和基于高斯基函数的卡尔曼滤波方法,还可以跟踪ERP的幅度趋势变异性。与基于高斯基函数的卡尔曼滤波方法相比,所提方法降低了计算量。真实脑电ERP提取实验表明本方法较好地从单次记录中提取出了事件相关电位,并可解释ERP因适应和应激引起的趋势变异性。     

结合子空间方法和提升小波变换提取事件相关脑电位

我们提出了一种提取事件相关脑电位的复合方法。它用奇异值分解方法将含噪信号分解为噪声子空间和信号子空间,将含噪信号正交投影到信号子空间进行初步除噪,随后将得到的信号进行提升小波变换,对变换结果进行一维小波重构进一步去除噪声,最后提取出ERP成分。介绍了基于奇异值分解的子空间方法和对信号进行提升小波去噪的实现方法。仿真结果表明,结合两种方法提取事件相关脑电信号时,比单独采用其中一种方法的效果要好,并可减少提取事件相关电位所需的实验次数。对实验数据的处理结果表明,该方法的实际处理效果良好。     

基于自适应中值滤波的脑事件关联电位单次提取

提出一种基于自适应中值滤波的脑事件关联电位单次提取方法。刺激方式为odd-ball模式，直接用实测的数据进行实验。背景噪声从波形上看呈现为脉冲状形式，中值滤波可以有效地去除脉冲噪声。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中值滤波器处理，最后对结果进行低通平滑。实验结果表明，本文提出的方法不但具有良好的单次提取效果，而且具有很强的适应性和稳健性。     

基于径向基函数神经网络的多电极阵列信号脉冲电位的分类

通过径向基函数神经网络的分析,对神经元脉冲电位信号提出了新的分类方法。对原始信号进行峰电位检测,获得脉冲电位信号样本,以主成分进行预分类,选取与类中心方差小的典型脉冲电位集合作为径向基网络的训练样本,让神经网络进行自适应学习,以实现对原始信号的分类。仿真结果表明,在对模拟的脉冲电位信号进行分类时此方法的错误率比主成分聚类法和形状聚类法小。多电极细胞外记录的海马神经元细胞电活动信号应用此方法分类也取得了较好的效果。     

基于自适应移动平均的脑事件关联电位单次提取

提出一种基于自适应移动平均的脑事件关联电位单次提取方法。采用零交叉方法对背景噪声进行分析,自适应提取单次记录中的最大脉冲噪声宽度,将其作为窗口大小的基准,然后用移动平均算法对测量信号进行处理。实验结果表明,我们提出的方法不但具有良好的单次提取效果,而且具有较强的适应性和稳健性。     

听觉诱发电位提取方法的研究与分析

论述了运用小波变换进行听觉诱发电佗单次提取的原理、方法和实验分析。结果表明，对单次试验信号，经小波变换及相关分析后，可从带自发脑电干扰的信号中提取诱发电位信号。小波变换分析方法与传统的叠加平均方法相比，可减少试验次数，缩短检测周期。     

诱发电位的非线性动态提取方法

以不同的潜伏期和幅度非线性迭加在自发脑电上的单次诱发电位，是诱发电位和自发脑电的一种可能的非线性组合。对诱发电位进行非线性动态提取是近年来备受关注而又十分困难的课题。本综述了三种非线性的诱发电位动态提取方法：自适应径向基函数神经网络方法、小波变换的非线性阈值系数方法以及模糊聚类方法，中着重介绍了以上方法的基本思路和典型应用实例。     

脑干听觉诱发电位的非线性动态提取

临床上常用的平均脑干听觉诱发电位（Brainstem Auditory Evoked Potential，BAEP）无法描述脑干功能的动态特性，从背景噪声中单次或少次动态提取的BAEP才是反映脑干功能的理想信号。径向基函数神经网络（Radial Basis Function Neural Network，RBFNN）已被用于BAEP的非线性动态提取，但是对于“淹没”在噪声中的信噪比很小的BAEP提取效果不好。本研究用移动窗口平均（MovingWindowAverage，MWA）先对含噪声的BAEP进行动态少次平均提高信噪比，然后再用RBFNN进行BAEP的非线性提取，在保留了绝大部分BAEP动态信息的前提下改善了RBFNN的提取性能。为了验证方法的可行性，构建了信噪比为-25dB的仿真BAEP-噪声序列，经MWA和RBFNN动态提取后相对均方误差约为19％，比仅用RBFNN时误差降低了6％左右。将上述方法用于实际测取的BAEP，可以得到每个子波形和平均BAEP波形波幅趋势大体相同的动态序列，这个BAEP动态序列为应用非线性动力学研究脑干功能动态特性打下了基础。     

A tracing evoked potential estimator

The paper presents an adaptive Gaussian radial basis function neural network (RBFNN) for rapid estimation of evoked potential (EP). Usually, a recorded EP is severely contaminated by background ongoing activities of the brain. Many approaches have been reported to enhance the signal-to-noise ratio (SNR) of the recorded signal. However, non-linear methods are seldom explored due to their complexity and the fact that the non-linear characteristics of the signal are generally hard to determine. An RBFNN possesses built-in non-linear activation functions that enable the neural network to learn any function mapping. An RBFNN was carefully designed to model the EP signal. It has the advantage of being linear-in-parameter, thus a conventional adaptive method can efficiently estimate its parameters. The proposed algorithm is simple so that its convergence behaviour and performance in signal-to-noise ratio (SNR) improvement can be mathematically derived. A series of experiments carried out on simulated and human test responses confirmed the superior performance of the method. In a simulation experiment, an RBFNN having 15 hidden nodes was trained to approximate human visual EP (VEP). For detecting gene rate=0.005) speeded up the estimation remarkably by using only 80 ensembles to achieve a result comparable to that obtained by averaging 1000 ensembles.     

A tracing evoked potential estimator

The paper presents an adaptive Gaussian radial basis function neural network (RBFNN) for rapid estimation of evoked potential (EP). Usually, a recorded EP is severely contaminated by background ongoing activities of the brain. Many approaches have been reported to enhance the signal-to-noise ratio (SNR) of the recorded signal. However, non-linear methods are seldom explored due to their complexity and the fact that the non-linear characteristics of the signal are generally hard to determine. An RBFNN possesses built-in non-linear activation functions that enable the neural network to learn any function mapping. An RBFNN was carefully designed to model the EP signal. It has the advantage of being linear-in-parameter, thus a conventional adaptive method can efficiently estimate its parameters. The proposed algorithm is simple so that its convergence behaviour and performance in signal-to-noise ratio (SNR) improvement can be mathematically derived. A series of experiments carried out on simulated and human test responses confirmed the superior performance of the method. In a simulation experiment, an RBFNN having 15 hidden nodes was trained to approximate human visual EP (VEP). For detecting human brain stem auditory EP (BAEP), the approach (40 hidden nodes and convergence rate = 0.005) speeded up the estimation remarkably by using only 80 ensembles to achieve a result comparable to that obtained by averaging 1000 ensembles.     

Reconstruction of gastric slow wave from finger photoplethysmographic signal using radial basis function neural network

Extraction of extra-cardiac information from photoplethysmography (PPG) signal is a challenging research problem with significant clinical applications. In this study, radial basis function neural network (RBFNN) is used to reconstruct the gastric myoelectric activity (GMA) slow wave from finger PPG signal. Finger PPG and GMA (measured using Electrogastrogram, EGG) signals were acquired simultaneously at the sampling rate of 100 Hz from ten healthy subjects. Discrete wavelet transform (DWT) was used to extract slow wave (0–0.1953 Hz) component from the finger PPG signal; this slow wave PPG was used to reconstruct EGG. A RBFNN is trained on signals obtained from six subjects in both fasting and postprandial conditions. The trained network is tested on data obtained from the remaining four subjects. In the earlier study, we have shown the presence of GMA information in finger PPG signal using DWT and cross-correlation method. In this study, we explicitly reconstruct gastric slow wave from finger PPG signal by the proposed RBFNN-based method. It was found that the network-reconstructed slow wave provided significantly higher (P < 0.0001) correlation (≥0.9) with the subject’s EGG slow wave than the correlation obtained (≈0.7) between the PPG slow wave from DWT and the EEG slow wave. Our results showed that a simple finger PPG signal can be used to reconstruct gastric slow wave using RBFNN method.     

Modelling the time-keeping function of the central pattern generator for locomotion using artificial sequential neural network

The paper investigates the ability of a sequential neural network to model the time-keeping function (fundamental frequency oscillation) of a central pattern generator for locomotion. The intention is not to strive for biological fidelity, but rather to ensure that the network obeys the organisational and operational principles of central pattern generators developed through empirical research. The timing function serves to produce the underlying locomotor rhythm which can be transformed by nonlinear static shaping functions to construct the necessary locomotor activation patterns. Using two levels of tonic activations in the form of a step increase, a network consisting of nine processing units was successfully trained to output both sine and cosine waveforms, whose frequencies were modified in response to the level of input activation. The network's ability to generalise was demonstrated by appropriately scaling the frequency of oscillation in response to a range of input amplitudes, both within and outside the values on which it was trained. A notable and fortunate result was the model's failure to oscillate in the absence of input activation, which is a necessary property of the CPG model. It was further demonstrated that the oscillation frequency of the output waveforms exhibited both a high temporal stability and a very low sensitivity to input noise. The results indicate that the sequential neural network is a suitable candidate to model the time-keeping functions of the central pattern generator for locomotion.     

Prediction of Seizure Onset in an <Emphasis Type="Italic">In-Vitro</Emphasis> Hippocampal Slice Model of Epilepsy Using Gaussian-Based and Wavelet-Based Artificial Neural Networks

We propose that artificial neural networks (ANNs) can be used to predict seizure onsets in an in-vitro hippocampal slice model capable of generating spontaneous seizure-like events (SLEs) in their extracellular field recordings. This paper assesses the effectiveness of two ANN prediction schemes: Gaussian-based artificial neural network (GANN) and wavelet-based artificial neural network (WANN). The GANN prediction system consists of a recurrent network having Gaussian radial basis function (RBF) nonlinearities capable of extracting the estimated manifold of the system. It is able to classify the underlying dynamics of spontaneous in-vitro activities into interictal, preictal and ictal modes. It is also able to successfully predict the onsets of SLEs as early as 60 s before. Improvements can be made to the overall seizure predictor design by incorporating time-varying frequency information. Consequently, the idea of WANN is considered. The WANN design entails the assumption that frequency variations in the extracellular field recordings can be used to compute the times at which onsets of SLEs are most likely to occur in the future. Progressions of different frequency components can be captured by the ANN using appropriate frequency band adjustments via pruning, after the initial wavelet transforms. In the off-line processing comprised of 102 spontaneous SLEs generated from 14 in-vitro rat hippocampal slices, with half of them used for training and the other half for testing, the WANN is able to predict the forecoming ictal onsets as early as 2 min prior to SLEs with over 75% accuracy within a 30 s precision window.     

Response planning in word typing: Evidence for inhibition

Typing is a pervasive phenomenon, yet the underlying neural processes have hardly been studied. Here, the mechanisms of keystroke preparation were studied with a typed picture‐naming task performed by expert typists. Electroencephalographic activities recorded over sensorimotor areas prior to first‐keystroke onset were examined with time‐frequency and event‐related potential (ERP) analyses. In the time‐frequency domain, a beta event‐related desynchronization was present bilaterally. In the ERP analyses, the activity was asymmetric, with negativity and positivity patterns developing over, respectively, contra‐ and ipsilateral recording sites. This pattern is similar to that observed in choice reaction time tasks, and thus can be interpreted as evidence of contralateral motor cortex activation accompanied by inhibition of the ipsilateral motor cortex. These data constitute the first electrophysiological demonstration of inhibitory activity in typing and pave the way to a thorough study of typing.     

Design of electrodes and current limits for low frequency electrical impedance tomography of the brain

For the novel application of recording of resistivity changes related to neuronal depolarization in the brain with electrical impedance tomography, optimal recording is with applied currents below 100 Hz, which might cause neural stimulation of skin or underlying brain. The purpose of this work was to develop a method for application of low frequency currents to the scalp, which delivered the maximum current without significant stimulation of skin or underlying brain. We propose a recessed electrode design which enabled current injection with an acceptable skin sensation to be increased from 100 μA using EEG electrodes, to 1 mA in 16 normal volunteers. The effect of current delivered to the brain was assessed with an anatomically realistic finite element model of the adult head. The modelled peak cerebral current density was 0.3 A/m², which was 5 to 25-fold less than the threshold for stimulation of the brain estimated from literature review.