基于多通道信息融合分析的表面肌电信号分解研究

通过多通道信息检测与融合分析来探讨表面肌电(sEMG)信号分解问题,以获取准确的运动单位动作电位(MUAP)模式判别。采用结合连续小波变换和假设检验的波形检测方式从多通道sEMG信号中提取动作电位波形,在对动作电位波形空间分布特征信息融合分析的基础上通过层次聚类方法来确定MUAP类别数目,再利用模糊k均值算法以及针对未归类波形的波形剥离方法实现多通道sEMG信号的准确分解。实验结果表明,多通道sEMG信号中MUAP信息得到有效检测和模式分类。所采用方法利用多通道sEMG信号细致地获取了MUAP波形空间分布信息,能够取得满意的分解效果。     

基于神经网络和递归模板对准技术的表面肌电信号分解

为了提高表面肌电信号(surface electromyography, sEMG)分解的准确率,我们利用空间相邻两通道sEMG信号的信息,采用联合低频小波分解系数作为运动单位动作电位(motor unit action potential, MUAP)活动段的特征,并将自组织特征映射(self-organizing feature map, SOFM)与学习向量量化(learning vector quantization, LVQ)网络结合起来,完成对MUAP波形的分类.同时为了实现对sEMG信号分解的完整性,采用一种基于递归的模板对准技术分解叠加波形.仿真信号和真实信号的实验表明,本方法具有较高的分解准确率,对于中低收缩力度下sEMG信号的分解十分有效.     

利用运动单位发放信息建立多通道信息的表面肌电信号分解

目的基于多通道信息的表面肌电(surfaceelectromyographic,sEMG)信号分解有助于弥补单通道分解时空间和发放信息不足的缺点.本文提出利用运动单位(motorunit,MU)的发放信息建立多通道 sEMG信号中属于同一 MU的模板映射关系,实现多导信号的信息互补,从而提高分解的准确率.方法对四导仿真信号先分别进行单通道分解,然后利用各通道之间的发放信息建立模板映射关系进行多通道分解.结果仿真实验结果显示单通道分解准确率平均为75%,多通道分解准确率为88%,表明利用MU发放信息建立模板映射关系进行sEMG信号分解能够提高分解有效性.结论将该方法应用于真实信号分解,也能有效得到 MU的波形和发放信息.     

基于三级滤波器的表面肌电信号降噪处理

表面肌电信号（surface electromyography,sEMG）是一种非平稳微弱信号,而它的低信噪比是造成对其进行分解十分困难的主要原因之一.本文针对sEMG信号的噪声特点,提出基于经验模态分解（empirical mode decomposition,EMD）的三级滤波器技术来对sEMG信号进行预处理,即采用频谱插值法去除工频干扰,采用形态学运算去除基线漂移,采用经验模态分解去除白噪声.实验结果表明,本文所提出的方法不仅能够提高sEMG信号的信噪比,也能有效地保留运动单位动作电位（motor unit action potential,MUAP）的波形信息,这将有利于对MUAP的识别从而提高对sEMG信号的分解准确率.     

基于非线性能量算子的表面肌电信号MUAP发放检测

采用基于平滑非线性能量算子(smoothed nonlinear energy operator, SNEO)的方法对表面肌电(surface EMG, SEMG)信号运动单位动作电位(motor unit action potential, MUAP)的发放信息进行检测,提出一种能较精确确定MUAP发放数目的阈值检测方法.利用这些方法分别对肌肉轻度收缩和中度收缩时的SEMG信号进行了MUAP发放检测实验,结果表明,对于轻度收缩时的SEMG信号,本文的方法十分有效,而对中度收缩时的SEMG信号也能获得比较满意的检测结果.     

模型生理参数对表面肌电信号影响的仿真研究

通过模型研究肌肉生理参数对表面肌电信号的影响。根据肌肉的形态结构和生理特征,从肌电信号的信号源-细胞内动作电位开始,仿真了单肌纤维动作电位,由此合成了运动单位动作电位,再利用运动单位的募集发放模型,进一步仿真了运动单位动作电位序列,并最终完成了对表面肌电信号的仿真。在此基础上研究了极化区域宽度、跨膜电流密度分布和肌肉组织各向异性3个重要的模型生理参数对表面肌电信号统计特征的影响,得到了一些有价值的结果。实验结果表明,仿真肌电信号能够有效表征肌肉电生理变化过程。     

基于HHT变换的SEMG信号中MUAP的数目估计

对表面肌电（SEMG）信号中单位动作电位（MUAP）的数目进行估计可为神经肌肉控制的理论研究和神经肌肉疾病的诊断开辟新途径,本文给出了一种基于Hilbert-Huang变换（HHT）的表面肌电信号中运动MUAP数目估计方法.通过对SEMG信号经验模态分解后的第一内禀模态函数分量进行瞬时频率分析,利用其瞬时频率极值点的计数即可估计出运动MUAP数目.仿真信号与真实信号的实验结果均表明,基于HHT的SEMG信号中MUAP的估计方法是有效的.     

具有迭加动作电位波形的EMG信号自动分解研究

本文提出一种不需人工干预的ＥＭＧ信号自动分解算法。算法的学习阶段使用初始的一小段ＥＭＧ记录以估计不同运动单位的动作电位模板。分解阶段根据运动单位发放统计特性，采有基于最小错误率的贝叶斯准则，将ＥＭＧ信号分解为构成它的各运动单位动作电位序列，这里重点讨论了具有迭加动作电位波形的ＥＭＧ信号分解问题。最后利用合成的模拟ＥＭＧ信号及真实ＥＭＧ信号对算法进行了检验，表明其具有很高的分解正确率     

基于先验模板的表面肌电信号渐进分解算法研究

本研究通过将表面肌电信号（sEMG）分解为运动单元动作电位序列（MUAPTs）,来研究神经 肌肉系统中运动单元（MU）的募集与发放模式。针对高收缩力情况下MUAP叠加问题,首先采用FastICA算法和小波包去噪算法对信号进行预处理;然后基于先验知识构建了4种形态的可伸缩MUAP模板;最后,采用“先大后小”的渐进识别方式,逐个对MUAP进行自动提取。在此基础上,还将该算法应用于8名受试者（3组/人）不同手指活动模式下的指浅屈肌多通道（12通道）sEMG分解;单通道分解结果显示,高力量水平下sEMG中的主体MUAPTs能够被有效检测和分类;统计结果证实,随着力量水平的增加,MUAP的数目增加;不同大小MUAP比重的变化与活动手指和力量水平具有显著的相关性。本文的实验结果,初步验证了利用先验模板从sEMG中渐进提取MUAP的可行性,为sEMG分解和进一步研究MU发放规律提供于一种新的思路。     

食指力量对指浅屈肌运动单元募集模式的影响

目的检测不同食指力量水平下指浅屈肌运动单元的募集模式。方法设计食指20%、40%、60%最大随意收缩力量(maximum voluntary contraction,MVC)3个单指力量输出任务,采用8×1(行×列)阵列电极采集8名受试者的指浅屈肌sEMG信号,利用快速独立分量分析算法提取sEMG信号中运动单位动作电位(motor unit action potential,MUAP)信息,分析不同类型MUAP发放模式与力量的相关性。结果在原始信号中成功提取4种MUPA,且随力量水平的增加,MUAP总数目呈现递增趋势;不同力量水平下,4种类型MUAP所占比重不同,且随力量变化趋势不同。结论不同力量水平下,指浅屈肌改变运动单元募集模式以产生相应肌力。     

A Study on sEMG Simulation Modeling and Its Decomposition Methods

In this paper, we establish a surface electromyography(sEMG) signal model and study the signal decomposition method from noisy background. Firstly, single fiber action potential (SFAP), motor unit action potential (MUAP) and motor unit action potential train(MUAPT) are simulated based on the tripolar signal source model, and then the sEMG is obtained; secondly, the simulated sEMG signal is extracted from the mixed signals that consists of white noises, power frequency interference signal and electrocardio signal by independent component analysis (ICA) algorithms; lastly, the spikes corresponding to each motor unit action potential from the simulated sEMG signals were detected by applying the wavelet transform (WT) method. Simulation results showed that sEMG model could describe the physiological process of sEMG, ICA and WT methods could extract the sEMG signal and its features, which will lay a foundation for further classifying the MUAP.     

利用自组织竞争神经网络提取NEMG信号的MUAP模板

采用自组织竞争人工神经网络,完成对针电极肌电信号（NEMG）的运动单位动作电位（MUAP）的模式分类。MUAP波形的特征取自于其自回归（AR）模型系数a1～ap及激励白噪的功率ε^p构成的特征向量。模拟NEMG信号和真实NEMG信号的实验结果表明,这种分类方法具有很高的正确,从而为NEMG信号分解研究中提取MUAP模板提供了一条新的途径。     

Resolving superimposed motor unit action potentials

A new algorithm to resolve superimposed motor unit action potentials (MUAPs) is described, which uses a reduced search space and is based on the peel off approach. Knowledge specific to the problem domain, such as temporal relationships between and within motor unit action potential trains and MUAP energy information, is used to reduce the search space of motor units, possibly contributing to a superposition. The algorithm is tested using real electromyographic signals, and it demonstrates robust performance across the signals tested. For the signals tested, the average total resolution rate is 94%, the average correct resolution rate is 99.2% and the average error rate is 0.85%.     

Model for decomposition of the motor unit action potential 1 The algorithm

A model for decomposition of the motor unit action potential (MUAP) into its constituent single-fibre action potentials is presented. It finds an optimal fit of a set of simulated single-fibre action potentials (SSFAPs) to the MUAP. The SSFAPs are assumed to originate from muscle fibres at different distances from the electrode, having various delays in time. Two methods for decomposition of the MUAP are derived from this model: first, that the MUAP is decomposed into a fixed set of SSFAPs; and secondly that the MUAP is decomposed into an adaptive, expanding set of SSFAPs. In the second method three steps are used repeatedly. First, the MUAP is cross-correlated with a collection of four SSFAPs. Then the most similar SSFAPs are used to reconstruct the original MUAP. The reconstruction thus obtained is subtracted from the original MUAP to detect activity not yet imitated. This difference (‘residual’) is again used for cross-correlation, restarting in step 1. After a suitable number of iterations, the MUAP is optimally imitated by a set of SSFAPs. The set of SSFAPs, obtained as described, is assumed to give information about underlying anatomical and physiological data (such as fibre number, fibre density, impulse dispersion) of the motor unit under study.     

Modified motor unit number index: A simulation study of the first dorsal interosseous muscle

The motor unit number index (MUNIX) technique has provided a quick and convenient approach to estimating motor unit population changes in a muscle. Reduction in motor unit action potential (MUAP) amplitude can lead to underestimation of motor unit numbers using the standard MUNIX technique. This study aims to overcome this limitation by developing a modified MUNIX (mMUNIX) technique. The mMUNIX uses a variable that is associated with the area of compound muscle action potential (CMAP) rather than an arbitrary fixed value (20 mV ms) as used in the standard MUNIX to define the output. The performance of the mMUNIX was evaluated using motoneuron pool and surface electromyography (EMG) models. With a fixed motor unit number, the mMUNIX output remained relatively constant with varying degrees of MUAP amplitude changes, while the standard MUNIX substantially underestimated the motor unit number in such cases. However, when MUAP amplitude remained unchanged, the mMUNIX showed less sensitivity than the standard MUNIX in tracking motor unit loss. The current simulation study demonstrated both the advantages and limitations of the standard and modified MUNIX techniques, which can help guide appropriate application and interpretation of MUNIX measurements.     

Adaptive spatial filtering of multichannel surface electromyogram signals

Spatial filtering of surface electromyography (EMG) signal can be used to enhance single motor unit action potentials (MUAPs). Traditional spatial filters for surface EMG do not take into consideration that some electrodes could have poor skin contact. In contrast to the traditional a priori defined filters, this study introduces an adaptive spatial filtering method that adapts to the signal characteristics. The adaptive filter, the maximum kurtosis filter (MKF), was obtained by using the linear combination of surrounding channels that maximises kurtosis. The MKF and conventional filters were applied to simulated EMG signals and to real EMG signals recorded with an electrode grid to evaluate their performance in detecting single motor units. The MKF was compared with conventional spatial filtering methods. Simulated signals, with different levels of spatially correlated noise, were used for comparison. The influence of one electrode with poor skin contact was also investigated. The MKF was found to be considerably better at enhancing a single MUAP than conventional methods for all levels of spatial correlation of the noise. For a spatial correlation of 0.97 of the noise, the improvement in the signal-to-noise ratio, where a MUAP could be detected, was at least 6 dB. With a simulated poor skin contact for one electrode, the improvement over the other methods was at least 19 dB.     

A new and fast approach towards sEMG decomposition

The decomposition of high-density surface EMG (HD-sEMG) interference patterns into the contribution of motor units is still a challenging task. We introduce a new, fast solution to this problem. The method uses a data-driven approach for selecting a set of electrodes to enable discrimination of present motor unit action potentials (MUAPs). Then, using shapes detected on these channels, the hierarchical clustering algorithm as reported by Quian Quiroga et al. (Neural Comput 16:1661–1687, 2004) is extended for multichannel data in order to obtain the motor unit action potential (MUAP) signatures. After this first step, more motor unit firings are obtained using the extracted signatures by a novel demixing technique. In this demixing stage, we propose a time-efficient solution for the general convolutive system that models the motor unit firings on the HD-sEMG grid. We constrain this system by using the extracted signatures as prior knowledge and reconstruct the firing patterns in a computationally efficient way. The algorithm performance is successfully verified on simulated data containing up to 20 different MUAP signatures. Moreover, we tested the method on real low contraction recordings from the lateral vastus leg muscle by comparing the algorithm’s output to the results obtained by manual analysis of the data from two independent trained operators. The proposed method showed to perform about equally successful as the operators.     

Model for decomposition of the motor unit action potential 2 Analysis of simulated motor unit action potentials

A model for decomposition of the motor unit action potential (MUAP) which finds an optimal fit of a set of simulated single-fibre action potentials (SSFAPs) to the original MUAP is tested. The composition of SSFAPs which best produces the MUAP is assumed to carry information about the actual distribution of single-fibre action potentials generating the MUAP. Two methods are derived from the model. The first makes use of a fixed set of SSFAPs. In the second method, a gradually expanding set of SSFAPs is built, using a sequence of crosscorrelation, optimal reconstruction and subtraction. In the paper MUAPs are constructed under various well defined conditions. The MUAPs are decomposed by the two methods, and the results are compared with traditional MUAP parameters. Under these conditions, the model obtains parameters with closer biological connections compared with traditional measures.