Analysis and classification of compressed EMG signals by wavelet transform via alternative neural networks algorithms

We propose intelligent methods for classifying three different muscle types, i.e. biceps, frontallis and abductor pollicis brevis muscles, with low computational complexity. For this aim, electromyogram (EMG) signals are recorded and modelled by using an auto-regressive (AR) model. As the size of the EMG signals is usually large, the computational complexity of artificial neural network (ANN) systems drastically increases. Therefore, in the proposed scheme EMG signals are pre-processed by using a wavelet transform and then they are modelled by employing an AR approach. The AR coefficients are used to train and test the ANNs. Experimental results show that the highest achieved classification accuracy is more than 95% in the case of EMG signals pre-processed by wavelet transform. The wavelet transform-based pre-processing significantly increases the performance rates compared to standard multilayer perceptron and general regression neural networks algorithms.     

The return of the vole cycle in southern Finland refutes the generality of the loss of cycles through ‘climatic forcing’

Multiannual cycles in the abundance of voles and other animals have been collapsing in the last decades. It has been proposed that this phenomenon is ‘climatically forced’ by milder winters. We here consider the dynamics of bank and field voles during more than two decades in two localities (170 km apart) in southern Finland. Using wavelet analysis, we show that a clear 3‐year cycle disappeared in the mid 1990s. However, the vole cycle returned in both localities after about 5 years despite winters becoming increasingly milder. In both localities, vole cycles were mainly determined by bank voles after the period of noncyclic dynamics, whereas field voles were dominant before this irregularity. Wavelet coherency analysis shows that spatial synchrony temporarily broke down during the period of noncyclic dynamics, but was fully restored afterwards. The return of the cycle despite ongoing rapid climate change argues against ‘climatic forcing’ as a general explanation for loss of cycles. Rather, the population‐dynamical consequences of climate change may be dependent on the local species composition and mechanism of delayed density dependence.     

大鼠慢波睡眠和睡眠过渡期脑电信号的小波变换分析

EEG信号经常包含许多快速的时变信息 ,将较长时间段的EEG信号近似看作平稳信号 ,进行FFT谱估计 ,存在其局限性。应用多分辨率小波变换方法 ,在频域和时域上可以同时定位分析大鼠慢波睡眠和睡眠过渡期脑电的动态变化特性。采用慢性埋植电极记录自由活动大鼠的皮层脑电 ,将信号用小波变换分解成δ、θ、α和 β四个分量 ,求各分量的功率和功率百分比的时间变化曲线 ,并与FFT功率谱分析结果进行比较。结果表明 :慢波睡眠期EEG中有 2 6 .2 %± 7.7%的时间段上δ分量功率小于总功率的 5 0 % ,且δ分量较大时 ,其他分量较小 ;δ分量较小时 ,其他分量较大 ,差别显著。此结果揭示了δ节律与θ和α节律之间的一种互补关系。而传统的FFT功率谱分析方法只能显示δ分量为主 (占总功率 70 .6 %± 6 .4 % )的功率谱 ,不能提供时变信息。对于睡眠过渡期的非稳态EEG信号 ,利用小波变换分解得到的θ和α分量可以鉴别出睡眠纺锤波 ,计算睡眠纺锤波的平均持续时间 ,并比较纺锤波和非纺锤波时期各个频谱分量的变化情况。由此可见 ,小波变换可用于计算新的EEG时频定量分析指标用于分析生理、病理和药理作用引起的睡眠EEG的变化过程 ,以弥补传统FFT功率谱分析的不足之处     

Fatigue-induced changes in muscle fiber action potentials estimated by wavelet analysis

We aimed to investigate fatigue-induced changes in the spectral parameters of slow (SMF) and fast fatigable muscle fiber (FMF) action potentials using discrete wavelet (DWT) and fast Fourier (FFT) transforms. Intracellular potentials were recorded during repetitive stimulation of isolated muscle fibers immersed in Ca²⁺-enriched medium, while extracellular potentials were obtained from muscle fibers pre-exposed to electromagnetic microwaves (MMW, 2.45 GHz, 20 mW/cm²). The changes in the frequency distribution of the action potentials during the period of uninterrupted fiber activity were used as criteria for fatigue assessment. The wavelet coefficients’ changes in the calculated frequency scales demonstrated a contribution of the increased [Ca²⁺]₀ to an earlier compression of the frequency spectrum towards lower ranges. Root mean square (RMS) analysis of the wavelet coefficients calculated from SMF potentials showed a reduction of the higher frequencies (scale 1) by 90% in elevated [Ca²⁺]₀ vs. 55% in controls and an increase of low frequencies (scale 5) by 323% vs. 187%, respectively. For FMF potentials a decrease of 71% vs. 59% for high frequencies (scale 1, elevated [Ca²⁺]₀ vs. control) and an increase of 386% vs. 295% in scale 5, respectively, were observed. MMW pre-exposure resulted in increased muscle fiber resistance to fatigue. The fatigue-induced decrease of potential high frequencies (SMF: 59% vs. 96%, MMW vs. control; FMF: 30% vs. 92%, respectively), and the increase of low frequencies (SMF: 200% vs. 207%, MMW vs. control; FMF: 93% vs. 314%, respectively) were significantly smaller and delayed in exposed muscle fibers. Data from RMS analysis indicate that DWT provides a reliable method for estimation of muscle fatigue onset and progression.     

Application of wavelet-based neural network on DNA microarray data

The advantage of using DNA microarray data when investigating human cancer gene expressions is its ability to generate enormous amount of information from a single assay in order to speed up the scientific evaluation process. The number of variables from the gene expression data coupled with comparably much less number of samples creates new challenges to scientists and statisticians. In particular, the problems include enormous degree of collinearity among genes expressions, likely violation of model assumptions as well as high level of noise with potential outliers. To deal with these problems, we propose a block wavelet shrinkage principal component (BWSPCA) analysis method to optimize the information during the noise reduction process. This paper firstly uses the National Cancer Institute database (NC160) as an illustration and shows a significant improvement in dimension reduction. Secondly we combine BWSPCA with an artificial neural network-based gene minimization strategy to establish a Block Wavelet-based Neural Network model in a robust and accurate cancer classification process (BWNN). Our extensive experiments on six public cancer datasets have shown that the method of BWNN for tumor classification performed well, especially on some difficult instances with large-class (more than two) expression data. This proposed method is extremely useful for data denoising and is competitiveness with respect to other methods such as BagBoost, RandomForest (RanFor), Support Vector Machines (SVM), K-Nearest Neighbor (KNN) and Artificial Neural Network (ANN).     

Fast Pseudo-Periodic Oscillation in the Rat Brain Voltage-gated Sodium Channel α Subunit

In the work reported here, we have investigated the changes in the activation and fast inactivation properties of the rat brain voltage-gated sodium channel (rNa_v 1.2a) α subunit, expressed heterologously in the Chinese Hamster Ovary (CHO) cells, by short depolarizing prepulses (10 – 1000 ms). The time constant of recovery from fast inactivation (τ_fast) and steady-state parameters for activation and inactivation varied in a pseudo-oscillatory fashion with the duration and amplitude of a sustained prepulse. A consistent oscillation was observed in most of the steady-state and non-inactivating current parameters with a time period close to 225 ms, although a faster oscillation of time period 125 ms was observed in the τ_fast. The studies on the non-inactivating current and steady-state activation indicate that the phase of oscillation varies from cell to cell. Co-expression of the β1 subunit with the α subunit channel suppressed the oscillation in the charge movement per single channel and free energy of steady-state inactivation, although the oscillation in the half steady-state inactivation potential remained unaltered. Incidentally, the frequencies of oscillation in the sodium channel parameters (4–8 Hz) correspond to the theta component of network oscillation. This fast pseudo-oscillatory mechanism, together with the slow pseudo-oscillatory mechanism found in these channels earlier, may contribute to the oscillations in the firing properties observed in various neuronal subtypes and many pathological conditions.     

Developing a multi-Kinect-system for monitoring in dairy cows: object recognition and surface analysis using wavelets

Camera-based systems in dairy cattle were intensively studied over the last years. Different from this study, single camera systems with a limited range of applications were presented, mostly using 2D cameras. This study presents current steps in the development of a camera system comprising multiple 3D cameras (six Microsoft Kinect cameras) for monitoring purposes in dairy cows. An early prototype was constructed, and alpha versions of software for recording, synchronizing, sorting and segmenting images and transforming the 3D data in a joint coordinate system have already been implemented. This study introduced the application of two-dimensional wavelet transforms as method for object recognition and surface analyses. The method was explained in detail, and four differently shaped wavelets were tested with respect to their reconstruction error concerning Kinect recorded depth maps from different camera positions. The images’ high frequency parts reconstructed from wavelet decompositions using the haar and the biorthogonal 1.5 wavelet were statistically analyzed with regard to the effects of image fore- or background and of cows’ or persons’ surface. Furthermore, binary classifiers based on the local high frequencies have been implemented to decide whether a pixel belongs to the image foreground and if it was located on a cow or a person. Classifiers distinguishing between image regions showed high (⩾0.8) values of Area Under reciever operation characteristic Curve (AUC). The classifications due to species showed maximal AUC values of 0.69.     

基于小波分析的大豆叶面积高光谱反演

实测了不同水肥耦合、经营制度及有效营养面积条件下的大豆(Glycinemax)冠层高光谱反射率与叶面积指数(LAI),并对光谱反射率、微分光谱与LAI的关系进行了分析;采用比值植被指数(RVI)与归一化植被指数(NDVI)建立了大豆LAI反演模型;采用小波分析对采集的光谱反射率数据进行了能量系数提取,并以小波能量系数作为自变量进行了单变量与多变量回归分析,对大豆LAI进行估算。结果表明:大豆LAI与光谱反射率在可见光波段呈负相关;在近红外波段呈正相关;微分光谱在红边处与大豆LAI密切相关(R2=0.92);RVI与NDVI可以提高大豆LAI的估算精度(R2分别达0.79、0.84);各植被指数各有优缺点,应根据需要进行选择;小波能量系数回归模型可以进一步提高大豆叶面积的估算水平,以一个特定小波能量系数作为自变量的回归模型,大豆LAI回归确定系数R2高达0.884;以4个和6个小波能量系数建立LAI回归分析模型(R2分别达0.92、0.93),2个模型LAI预测值与大豆LAI实测值线性回归确定性系数R2分别为0.90、0.92。比较可知,小波分析可以对高光谱进行特征变量提取,进而反演大豆生理参数,并且反演的LAI精度较光谱反射率、微分光谱及植被指数都有明显提高,小波分析在植被生理参数的高光谱提取方面有着广阔的应用前景。     

基于小波高频系数基因芯片数据的特征提取

结合小波分析理论与支持向量机理论,构造分类器模型,将前列腺癌基因芯片数据分成癌症和正常两种。本文着重研究小波高频系数基因芯片数据的特征提取,并通过实验对比小波高频系数和低频系数特征提取对分类器性能的影响。其中haar小波3层分解提取高频系数,送入分类器分类后,得到的正确分类率为93.31%。db1小波4层分解提取低频系数,送入分类器分类后,得到的正确分类率为93.53%。小波低频系数特征提取分类效果总体上好于高频系数,分类器性能稳定。