Research of fetal ECG extraction using wavelet analysis and adaptive filtering

Extracting clean fetal electrocardiogram (ECG) signals is very important in fetal monitoring. In this paper, we proposed a new method for fetal ECG extraction based on wavelet analysis, the least mean square (LMS) adaptive filtering algorithm, and the spatially selective noise filtration (SSNF) algorithm. First, abdominal signals and thoracic signals were processed by stationary wavelet transform (SWT), and the wavelet coefficients at each scale were obtained. For each scale, the detail coefficients were processed by the LMS algorithm. The coefficient of the abdominal signal was taken as the original input of the LMS adaptive filtering system, and the coefficient of the thoracic signal as the reference input. Then, correlations of the processed wavelet coefficients were computed. The threshold was set and noise components were removed with the SSNF algorithm. Finally, the processed wavelet coefficients were reconstructed by inverse SWT to obtain fetal ECG. Twenty cases of simulated data and 12 cases of clinical data were used. Experimental results showed that the proposed method outperforms the LMS algorithm: (1) it shows improvement in case of superposition R-peaks of fetal ECG and maternal ECG; (2) noise disturbance is eliminated by incorporating the SSNF algorithm and the extracted waveform is more stable; and (3) the performance is proven quantitatively by SNR calculation. The results indicated that the proposed algorithm can be used for extracting fetal ECG from abdominal signals.     

基于EMD的自适应ECG信号基线漂移消除方法研究

对比目前使用EMD或改进EMD方法进行的心电（ECG）信号基线漂移去除算法的实现。本文在详细考察EMD方法过程的基础上,提出一种与EMD物理意义高度契合的完全自适应的基线漂移算法,通过计算ECG平均心率周期,与EMD分解产生的IMF分量的“周期”进行对比,分离出不属于ECG信号的低频IMF分量,然后重构其余IMF分量得到去除基线漂移的ECG信号。使用美国麻省理工学院提供的MIT-BIH心率失常数据库中的原始ECG对本文提出的基线漂移去除方法进行定性分析。使用ECGSYN（实际ECG波形发生器）产生模拟干净的ECG信号,加入已知的低频信号作为基线漂移噪声,对本文提出的基线漂移去除方法进行定量分析。     

一种基于经验小波变换的心电信号室性早搏检测算法

针对心电信号中的室性早搏心拍检测问题,使用经验小波变换（EWT）实现心电信号的自适应分解。根据心电信号时频能量变化特征,提出了一种低复杂度的频域累积能量特征计算方法,并分析了室性早搏与正常心电信号的特征差异性。最后利用反向传播神经网络在MIT-BIH心电数据库上进行心拍样本训练与识别测试。结果表明基于EWT的特征提取避免了传统时域特征提取中的QRS波群检测过程,降低了其它干扰因素对诊断结果的影响,具有较高的分类精度与良好的鲁棒性,总体敏感度与总体阳性检测率分别达到96.55%和97.73%。     

ECG beat classification using empirical mode decomposition and mixture of features

Computer-aided analysis is useful in predicting arrhythmia conditions of the heart by analysing the recorded ECG signals. In this work, we proposed a method to detect, extract informative features to classify six types of heartbeat of ECG signals obtained from the MIT-BIH Arrhythmia database. The powerful discrete wavelet transform (DWT) is used to eliminate different sources of noises. Empirical mode decomposition (EMD) with adaptive thresholding has been used to detect precise R-peaks and QRS complex. The significant features consists of temporal, morphological and statistical were extracted from the processed ECG signals and combined to form a set of features. This feature set is classified with probabilistic neural network (PNN) and radial basis function neural network (RBF-NN) to recognise the arrhythmia beats. The process achieved better result with sensitivity of 99.96%, and positive predictivity of 99.81 with error rate of 0.23% in detecting the QRS complex. In class-oriented scheme, the arrhythmia conditions are classified with accuracy of 99.54%, 99.89% using PNN and RBF-NN classifier respectively. The obtained result confirms the superiority of the proposed scheme compared to other published results cited in literature.     

Removing ECG noise from surface EMG signals using adaptive filtering

Surface electromyograms (EMGs) are valuable in the pathophysiological study and clinical treatment for dystonia. These recordings are critically often contaminated by cardiac artefact. Our objective of this study was to evaluate the performance of an adaptive noise cancellation filter in removing electrocardiogram (ECG) interference from surface EMGs recorded from the trapezius muscles of patients with cervical dystonia. Performance of the proposed recursive-least-square adaptive filter was first quantified by coherence and signal-to-noise ratio measures in simulated noisy EMG signals. The influence of parameters such as the signal-to-noise ratio, forgetting factor, filter order and regularisation factor were assessed. Fast convergence of the recursive-least-square algorithm enabled the filter to track complex dystonic EMGs and effectively remove ECG noise. This adaptive filter procedure proved a reliable and efficient tool to remove ECG artefact from surface EMGs with mixed and varied patterns of transient, short and long lasting dystonic contractions.     

基于经验模态分解自适应滤波的胎儿心电信号提取

目的提出了一种基于经验模态分解自适应滤波的胎儿心电信号提取法。方法首先利用经验模态分解算法对孕妇腹部信号进行分解得到一组内模函数(IMF),然后将这组IMF作为自适应滤波器的主输入信号,并将孕妇胸部信号作为参考输入信号。通过学习算法自适应组合IMF,滤除母体心电信号成分,从而提取胎儿心电信号。结果与结论基于仿真和临床的实验结果表明,该方法提取的胎儿心电信号误差小,性能优于传统的最小均方和归一化最小均方自适应滤波算法。     

Electrocardiogram signals de-noising using lifting-based discrete wavelet transform

This paper introduces an effective technique for the denoising of electrocardiogram (ECG) signals corrupted by nonstationary noises. The technique is based on a second generation wavelet transform and level-dependent threshold estimator. Here, wavelet coefficients of ECG signals were obtained with lifting-based wavelet filters. A lifting scheme is used to construct second-generation wavelets and is an alternative and faster algorithm for a classical wavelet transform. The overall denoising performance of our proposed method is considered in relation to several measuring parameters, including types of wavelet filters (Haar, Daubechies 4 (DB4), Daubechies 6 (DB6), Filter(9-7), and Cubic B-splines), thresholding method, and decomposition depth. Three different kinds of noise were considered in this work: muscle artifact noise, electrode motion artifact noise, and white noise. Global performance is evaluated by means of the signal-to-noise ratio and visual inspection. Numerical results comparing the performance of the proposed method with that of nonlinear filtering techniques (median filter) are given. The results demonstrate consistently superior denoising performance of the proposed method over median filtering.     

A hybrid ECG compression algorithm based on singular value decomposition and discrete wavelet transform

Increasing use of computerized ECG processing systems requires effective electrocardiogram (ECG) data compression techniques which aim to enlarge storage capacity and improve data transmission over phone and internet lines. This paper presents a compression technique for ECG signals using the singular value decomposition (SVD) combined with discrete wavelet transform (DWT). The central idea is to transform the ECG signal to a rectangular matrix, compute the SVD, and then discard small singular values of the matrix. The resulting compressed matrix is wavelet transformed, thresholded and coded to increase the compression ratio. The number of singular values and the threshold level adopted are based on the percentage root mean square difference (PRD) and the compression ratio required. The technique has been tested on ECG signals obtained from MIT-BIH arrhythmia database. The results showed that data reduction with high signal fidelity can thus be achieved with average data compression ratio of 25.2:1 and average PRD of 3.14. Comparison between the obtained results and recently published results show that the proposed technique gives better performance.     

ECG compression using wavelet transform and particle swarm optimization

A new adaptive thresholding mechanism to determine the significant wavelet coefficients of an electrocardiogram (ECG) signal is proposed. It is based on estimating thresholds for different sub-bands using the concept of energy packing efficiency (EPE). Then thresholds are optimized using the particle swarm optimization (PSO) algorithm to achieve a target compression ratio with minimum distortion. Simulation results on several records taken from the MIT-BIH Arrhythmia database show that the PSO converges exactly to the target compression after four iterations while the cost function achieved its minimum value after six iterations. Compared to previously published schemes, lower distortions are achieved for the same compression ratios.     

An optimized design of seizure detection system using joint feature extraction of multichannel EEG signals

The detection of seizure onset and events using electroencephalogram (EEG) signals are important tasks in epilepsy research. The literature available on seizure detection has discussed the implementation of advanced signal processing algorithms using tools accessed over the cloud. However, seizure monitoring application needs near sensor processing due to privacy and latency issues. In this paper, a real time seizure detection system has been implemented using an embedded system. The proposed system is based on ensemble empirical mode decomposition (EEMD) and tunable-Q wavelet transform (TQWT) algorithms. The analysis and classification of non-stationary EEG signals require the wavelet transform with high Q-factor. However, direct use of TQWT increases the computational complexity of feature extraction from multivariate EEG signals. In this paper, the first step is to process the signal by using EEMD to obtain 8 intrinsic mode functions (IMFs). The Kraskov (KraEn), sample (SampEn), and permutation (PermEn) entropy features of IMFs are extracted and based on optimum values, and 4 IMFs are decomposed using TQWT. Secondly, centered correntropy (CenCorrEn) features of the 1^st and 16^th sub-band of TQWT have been used as classifier inputs. The performance of multilayer perceptron neural networks (MLPNN), least squares support vector machine (LSSVM), and random forest (RF) classifiers has been tested on the multichannel EEG data recorded from a local hospital. The RF classifier has produced the highest accuracy of 96.2% in classifying the signals. The proposed scheme has been employed in developing an embedded seizure detection system to assist neurologists in making seizure diagnostic decisions.     

基于能量估计的小波阈值与经验模态分解相结合的滤波方法研究

本文针对基于经验模态分解(EMD)的时空滤波器存在的固有模态函数分量中频率混叠交叉,导致有用信号与噪声一起被滤除的问题,结合小波在时间、尺度两域表征信号局部特征的特性,提出了一种基于能量估计实现EMD分解层数确定,小波变换阈值处理与EMD相结合的时空滤波方法。该方法既利用小波变换多分辨率的特性,又结合EMD的自适应分解与希尔伯特(Hilbert)谱分析中瞬时频率与能量意义的关系,从而解决了有用信号在滤波时被削弱的问题。以MIT/BIH标准心电数据库数据为对象的实验结果表明,该方法对于生理信号这一类强噪声下的微弱信号是一种有效的数据处理方法。     

基于经验模态分解-小波包变换的表面肌电信号手势识别

为了提高表面肌电信号（sEMG）手部运动识别的正确率,比较常规的sEMG预处理和特征提取方法,提出一种基于经验模态分解（EMD）和小波包变换（WPT）的sEMG手势识别模型。首先,使用EMD方法将sEMG进行平稳化,得到一系列的固有模态函数。其次,求取各个固有模态函数与原始信号的相关性,选取相关性较高的前4个分量作为有效分量。然后,采用Db3小波函数进行WPT,提取小波包系数中的平均能量、平均绝对值、最大值、均方根和方差等特征。分别采用线性判别分析和支持向量机对12种手部运动进行模式识别。结果表明基于EMD和WPT的sEMG手势识别正确率比直接提取小波包系数中的特征识别正确率高。     

Spectral broadening of ophthalmic arterial Doppler signals using STFT and wavelet transform

In this study, short-time Fourier transform (STFT) and wavelet transform (WT) were used for spectral analysis of ophthalmic arterial Doppler signals. Using these spectral analysis methods, the variations in the shape of the Doppler spectra as a function of time were presented in the form of sonograms in order to obtain medical information. These sonograms were then used to compare the applied methods in terms of their frequency resolution and the effects in determination of spectral broadening in the presence of ophthalmic artery stenosis. A qualitative improvement in the appearance of the sonograms obtained using the WT over the STFT was noticeable. Despite the qualitative improvement in the individual sonograms, no quantitative advantage in using the WT over the STFT for the determination of spectral broadening index was obtained due to the poorer variance of the wavelet transform-based spectral broadening index and the additional computational requirements of the wavelet transform.     

Application of adaptive neuro-fuzzy inference system for epileptic seizure detection using wavelet feature extraction

Intelligent computing tools such as artificial neural network (ANN) and fuzzy logic approaches are demonstrated to be competent when applied individually to a variety of problems. Recently, there has been a growing interest in combining both these approaches, and as a result, neuro-fuzzy computing techniques have been evolved. In this study, a new approach based on an adaptive neuro-fuzzy inference system (ANFIS) was presented for epileptic seizure detection. The proposed ANFIS model combined the neural network adaptive capabilities and the fuzzy logic qualitative approach. Decision making was performed in two stages: feature extraction using the wavelet transform (WT) and the ANFIS trained with the backpropagation gradient descent method in combination with the least squares method. Some conclusions concerning the impacts of features on the detection of epileptic seizures were obtained through analysis of the ANFIS. The results are highly promising, and a comparative analysis suggests that the proposed modeling approach outperforms ANN model in terms of training performances and classification accuracies. The results confirmed that the proposed ANFIS model has some potential in epileptic seizure detection. The ANFIS model achieved accuracy rates which were higher than that of the stand-alone neural network model.     

Analysis of spike-wave discharges in rats using discrete wavelet transform

A feature is a distinctive or characteristic measurement, transform, structural component extracted from a segment of a pattern. Features are used to represent patterns with the goal of minimizing the loss of important information. The discrete wavelet transform (DWT) as a feature extraction method was used in representing the spike-wave discharges (SWDs) records of Wistar Albino Glaxo/Rijswijk (WAG/Rij) rats. The SWD records of WAG/Rij rats were decomposed into time-frequency representations using the DWT and the statistical features were calculated to depict their distribution. The obtained wavelet coefficients were used to identify characteristics of the signal that were not apparent from the original time domain signal. The present study demonstrates that the wavelet coefficients are useful in determining the dynamics in the time-frequency domain of SWD records.     

Investigating routes to chaos in the guinea-pig cochlea using the continuous wavelet transform and the short-time Fourier transform

The continuous wavelet transform (CWT) and the short-time Fourier transform (STFT) were used to analyze the time course of cellular motion in the guinea pig inner ear. The velocity responses of individual outer hair cells and Hensen's cells to amplitude modulated (AM) acoustical signals applied to the ear canal displayed characteristics typical of nonlinear systems, such as the generation of spectral components at harmonics of the carrier frequency. Nonlinear effects were particularly pronounced at the highest stimulus levels, where half-harmonic (and sometimes quarter-harmonic) components were also seen. The generation of these components was consistent with the behavior of a dynamical system entering chaos via a period-doubling route. A negative-stiffness Duffing oscillator model yielded period-doubling behavior similar to that of the experimental data. We compared the effectiveness of the CWT and the STFT for analyzing the responses to AM stimuli. The CWT (calculated using a high-Q Morlet-wavelet basis) and the STFT were both useful for identifying the various spectral components present in the AM velocity response of the cell. The high-Q Morlet wavelet CWT was particularly effective in distinguishing the lowest frequency components present in the response, since its frequency resolution is appreciably better than the STFT at low frequencies. Octave-band-based CWTs (using low-Q Morlet, Meyer, and Daubechies 4-tap wavelets) were largely ineffective in analyzing these signals, inasmuch as the frequency spacing between neighboring spectral components was far less than one octave.     

Artifact reduction in electrogastrogram based on empirical mode decomposition method

Severe contamination of the gastric signal in electrogastrogram (EGG) analysus by respiratory, motion, cardiac artifacts, and possible myoelectrical activity from other organs, poses a major challenge to EGG interpretation and analysis. A generally applicable method for removing a variety of artifacts from EGG recordings is proposed based on the empirical mode decomposition (EMD) method. This decomposition technique is adaptive, and appears to be uniquely suitable for nonlinear, non-stationary data analysis. The results show that this method, combined with instantaneous frequency analysis, effectively separate, identify and remove contamination from a wide variety of artifactual sources in EGG recordings.     

联合主成分分析-小波与整体平均经验模态分解的锋电位去噪方法

多通道微电极阵列记录的锋电位(Spike)十分微弱,极易受干扰,其含噪的特性影响了Spike检出的准确率。针对Spike检测过程中通常存在的独立白噪声、相关噪声与有色噪声,本文结合主成分分析(PCA)、小波分析和自适应时频分析,提出PCA-小波(PCAW)与整体平均经验模态分解(EEMD)联合的去噪新方法(PCWE)。首先,利用PCA提取多通道神经信号通道间的主成分作为相关噪声去除;然后利用小波阈值法对独立白噪声进行去除;最后利用EEMD把噪声分解到各层本质模态函数中,对有色噪声进行去除。仿真结果表明,PCWE使信噪比约提高2.67 dB,标准差约减小0.4μV,显著提高了Spike的检出精确率;实测数据结果表明,PCWE能使信噪比约提高1.33 dB,标准差约减小18.33μV,表现出良好的去噪性能。本文研究结果表明,PCWE可以提高Spike信号的可靠性,或可为神经信号的编码解码提供一种新型有效的锋电位去噪方法。     

Physical activities recognition from ambulatory ECG signals using neuro-fuzzy classifiers and support vector machines

Abstract

The use of wearable recorders for long-term monitoring of physiological parameters has increased in the last few years. The ambulatory electrocardiogram (A-ECG) signals of five healthy subjects with four body movements or physical activities (PA)—left arm up down, right arm up down, waist twisting and walking—have been recorded using a wearable ECG recorder. The classification of these four PAs has been performed using neuro-fuzzy classifier (NFC) and support vector machines (SVM). The PA classification is based on the distinct, time-frequency features of the extracted motion artifacts contained in recorded A-ECG signals. The motion artifacts in A-ECG signals have been separated first by the discrete wavelet transform (DWT) and the time–frequency features of these motion artifacts have then been extracted using the Gabor transform. The Gabor energy feature vectors have been fed to the NFC and SVM classifiers. Both the classifiers have achieved a PA classification accuracy of over 95% for all subjects.