Atrial activity extraction for atrial fibrillation analysis using blind source separation

This contribution addresses the extraction of atrial activity (AA) from real electrocardiogram (ECG) recordings of atrial fibrillation (AF). We show the appropriateness of independent component analysis (ICA) to tackle this biomedical challenge when regarded as a blind source separation (BSS) problem. ICA is a statistical tool able to reconstruct the unobservable independent sources of bioelectric activity which generate, through instantaneous linear mixing, a measurable set of signals. The three key hypothesis that make ICA applicable in the present scenario are discussed and validated: 1) AA and ventricular activity (VA) are generated by sources of independent bioelectric activity; 2) AA and VA present non-Gaussian distributions; and 3) the generation of the surface ECG potentials from the cardioelectric sources can be regarded as a narrow-band linear propagation process. To empirically endorse these claims, an ICA algorithm is applied to recordings from seven patients with persistent AF. We demonstrate that the AA source can be identified using a kurtosis-based reordering of the separated signals followed by spectral analysis of the sub-Gaussian sources. In contrast to traditional methods, the proposed BSS-based approach is able to obtain a unified AA signal by exploiting the atrial information present in every ECG lead, which results in an increased robustness with respect to electrode selection and placement.     

Temporally constrained ICA: an application to artifact rejection in electromagnetic brain signal analysis

Independent component analysis (ICA) is a technique which extracts statistically independent components from a set of measured signals. The technique enjoys numerous applications in biomedical signal analysis in the literature, especially in the analysis of electromagnetic (EM) brain signals. Standard implementations of ICA are restrictive mainly due to the square mixing assumption-for signal recordings which have large numbers of channels, the large number of resulting extracted sources makes the subsequent analysis laborious and highly subjective. There are many instances in neurophysiological analysis where there is strong a priori information about the signals being sought; temporally constrained ICA (cICA) can extract signals that are statistically independent, yet which are constrained to be similar to some reference signal which can incorporate such a priori information. We demonstrate this method on a synthetic dataset and on a number of artifactual waveforms identified in multichannel recordings of EEG and MEG. cICA repeatedly converges to the desired component within a few iterations and subjective analysis shows the waveforms to be of the expected morphologies and with realistic spatial distributions. This paper shows that cICA can be applied with great success to EM brain signal analysis, with an initial application in automating artifact extraction in EEG and MEG.     

Independent component approach to the analysis of EEG and MEG recordings

Multichannel recordings of the electromagnetic fields emerging from neural currents in the brain generate large amounts of data. Suitable feature extraction methods are, therefore, useful to facilitate the representation and interpretation of the data. Recently developed independent component analysis (ICA) has been shown to be an efficient tool for artifact identification and extraction from electroencephalographic (EEG) and magnetoencephalographic (MEG) recordings. In addition, ICA has been applied to the analysis of brain signals evoked by sensory stimuli. This paper reviews our recent results in this field.     

Imaging brain dynamics using independent component analysis

The analysis of electroencephalographic and magnetoencephalographic recordings is important both for basic brain research and for medical diagnosis and treatment. Independent component analysis (ICA) is an effective method for removing artifacts and separating sources of the brain signals from these recordings. A similar approach is proving useful for analyzing functional magnetic resonance brain imaging data. In this paper, we outline the assumptions underlying ICA and demonstrate its application to a variety of electrical and hemodynamic recordings from the human brain     

Enhancement of the ill-conditioned original recordings using novel ICA technique

The independent component analysis (ICA) method proposed in this study uses FastICA algorithm to improve the quality of the original recordings, which can be used as valuable pre-processing technique in signal processing methods. Initially, the ill-conditioned original audio recordings are separated using ICA methods and later, they are reconstructed using modified un-mixing matrix. The simulation results showed huge improvement of the original signal after reconstruction. The new method is found to be good because the accuracy is more compared to others in terms of the variance of the Gain matrix. The proposed method has potential applications in audio and biosignal processing techniques.     

Extracting Epileptic Feature Spikes Using Independent Component Analysis

In recent years, blind source separation （BSS） by independent component analysis （ICA） has been drawing much attention because of its potential applications in signal processing such as in speech recognition systems, telecommunication and medical signal processing. In this paper, two algorithms of independent component analysis （fixed-point IC,4 and natural gradient-flexible ICA） are adopted to extract human epileptic feature spikes from interferential signals. Experiment results show that epileptic spikes can be extracted from noise successfully. The kurtosis of the epileptic component signal separated is much better than that of other noisy signals. It shows that ICA is an effective tool to extract epileptic spikes from patients＇ electroencephalogram EEG and shows promising application to assist physicians to diagnose epilepsy and estimate the epileptogenic region in clinic.     

Spatiotemporal blind source separation approach to atrial activity estimation in atrial tachyarrhythmias

The analysis and characterization of atrial tachyarrhythmias requires, in a previous step, the extraction of the atrial activity (AA) free from ventricular activity and other artefacts. This contribution adopts the blind source separation (BSS) approach to AA estimation from multilead electrocardiograms (ECGs). Previously proposed BSS methods for AA extraction-e.g., independent component analysis (ICA)-exploit only the spatial diversity introduced by the multiple spatially-separated electrodes. However, AA typically shows certain degree of temporal correlation, with a narrowband spectrum featuring a main frequency peak around 3.5-9 Hz. Taking advantage of this observation, we put forward a novel two-step BSS-based technique which exploits both spatial and temporal information contained in the recorded ECG signals. The spatiotemporal BSS algorithm is validated on simulated and real ECGs from a significant number of atrial fibrillation (AF) and atrial flutter (AFL) episodes, and proves consistently superior to a spatial-only ICA method. In simulated ECGs, a new methodology for the synthetic generation of realistic AF episodes is proposed, which includes a judicious comparison between the known AA content and the estimated AA sources. Using this methodology, the ICA technique obtains correlation indexes of 0.751, whereas the proposed approach obtains a correlation of 0.830 and an error in the estimated signal reduced by a factor of 40%. In real ECG recordings, we propose to measure performance by the spectral concentration (SC) around the main frequency peak. The spatiotemporal algorithm outperforms the ICA method, obtaining a SC of 58.8% and 44.7%, respectively.     

Independent component analysis of high-density electromyography in muscle force estimation

Accurate force prediction from surface electromyography (EMG) forms an important methodological challenge in biomechanics and kinesiology. In a previous study (Staudenmann et al., 2006), we illustrated force estimates based on analyses lent from multivariate statistics. In particular, we showed the advantages of principal component analysis (PCA) on monopolar high-density EMG (HD-EMG) over conventional electrode configurations. In the present study, we further improve force estimates by exploiting the correlation structure of the HD-EMG via independent component analysis (ICA). HD-EMG from the triceps brachii muscle and the extension force of the elbow were measured in 11 subjects. The root mean square difference (RMSD) and correlation coefficients between predicted and measured force were determined. Relative to using the monopolar EMG data, PCA yielded a 40% reduction in RMSD. ICA yielded a significant further reduction of up to 13% RMSD. Since ICA improved the PCA-based estimates, the independent structure of EMG signals appears to contain relevant additional information for the prediction of muscle force from surface HD-EMG.     

Temporally constrained ICA-based foetal ECG separation

A new method for foetal ECG separation based on temporally constrained independent component analysis (ICA) is proposed. The maternal beat positions were regarded as a priori information and by applying constrained ICA, a maternal dominant signal could be extracted. Finally, a foetal ECG signal was obtained by the subtraction of this from a specific channel signal.     

Automatic identification and removal of scalp reference signal for intracranial EEGs based on independent component analysis

The pursuit of an inactive recording reference is one of the oldest technical problems in electroencephalography (EEG). Since commonly used cephalic references contaminate EEG and can lead to misinterpretation, extraction of the reference contribution is of fundamental interest. Here, we apply independent component analysis (ICA) to intracranial recordings and propose two methods to automatically identify and remove the reference based on the assumption that the scalp reference is independent from the local and distributed intracranial sources. This assumption, supported by our results, is generally valid because the reference scalp electrode is relatively electrically isolated from the intracranial electrodes by the skull's high resistivity. We point out that the linear model is underdetermined when the reference is considered as a source, and discuss one special underdetermined case for which a unique class of outputs can be separated. For this case most ICA algorithms can be applied, and we argue that intracranial or scalp EEGs follow this special case. We apply the two proposed methods to intracranial EEGs from three patients undergoing evaluation for epilepsy surgery, and compare the results to bipolar and average reference recordings. The proposed methods should have wide application in quantitative EEG studies.     

基于小波变换和盲信号分离的多通道肌电信号处理方法

 为了消除多通道表面肌电信号(SEMG)采集时形成的混叠现象,提出一种新的SEMG处理方法.该方法将小波变换和独立分量分析(ICA)结合,利用小波变换的去噪作用,滤除混合在原始SEMG中的部分噪声后作为ICA的输入信号,采用Infomax算法对输入信号实施盲分离,并引入相关系数验证ICA分量与源信号的一致性.实验结果表明,该方法用于多通道SEMG的盲分离是很有效的.     

FastICA算法在雷达快扫描中的应用

对于现代多普勒雷达来说,相关数据的计算精度依赖于估计单体内的独立样本数,独立样本数越多,精确度越大。为了增强雷达精确度和提高扫描速度,文中采用独立成分分析技术来增加独立样本数。独立成分分析算法通过迭代估计逐步逼近分离矩阵,从而得到所需的相对独立的因子。其中,快速不动点算法能在根本上显著提高计算方法的在线学习速度和可靠性,不仅收敛速度较快,而且收敛性不易破坏。利用基于极大似然估计的快速不动点算法先后对瑞利信号和模拟雷达信号进行了独立成分分析处理,并与常用的主成分分析技术进行了对比,获得了较好的试验结果,证实了独立成分分析技术的良好处理效果,为进一步应用于实际的多普勒雷达信号的处理奠定了基础。     

时变混合共轭梯度盲提取算法

针对传统独立分量分析(ICA)方法对时变信道跟踪能力较差的问题,提出了一种时变混合共轭梯度盲提取算法。该算法有效利用了各源信号的时序结构差异,仅利用其二阶统计量解决了具有不同功率谱密度的信号的分离,而无须估计信号的概率密度和计算高阶累积量,减少了运算的复杂度并可用于杂系信号混合的盲分离问题;同时,算法利用仅具有一个全局最优解的凸代价函数,采用计算简单并具有较好数值表现的自适应共轭梯度算法进行迭代,获得了更快的收敛速度和更好的稳定性能。仿真结果表明,该算法与传统ICA算法相比,具有对时变系统更好的跟踪能力。      

Artifact removal in magnetoencephalogram background activity with independent component analysis

The aim of this study was to assess whether independent component analysis (ICA) could be valuable to remove power line noise, cardiac, and ocular artifacts from magnetoencephalogram (MEG) background activity. The MEGs were recorded from 11 subjects with a 148-channel whole-head magnetometer. We used a statistical criterion to estimate the number of independent components. Then, a robust ICA algorithm decomposed the MEG epochs and several methods were applied to detect those artifacts. The whole process had been previously tested on synthetic data. We found that the line noise components could be easily detected by their frequency spectrum. In addition, the ocular artifacts could be identified by their frequency characteristics and scalp topography. Moreover, the cardiac artifact was better recognized by its skewness value than by its kurtosis one. Finally, the MEG signals were compared before and after artifact rejection to evaluate our method.     

Optimal reduction of MCG in fetal MEG recordings

Recording fetal magnetoencephalographic (fMEG) signals in-utero is a demanding task due to biological interference, especially maternal and fetal magnetocardiographic (MCG) signals. A method based on orthogonal projection of MCG signal space vectors (OP) was evaluated and compared with independent component analysis (ICA). The evaluation was based on MCG amplitude reduction and signal-to-noise ratio of fetal brain signals using exemplary datasets recorded during ongoing studies related to auditory evoked fields. The results indicate that the OP method is the preferable approach for attenuation of MCG and for preserving the fetal brain signals in fMEG recordings.     

Adaptive Step-Size Natural Gradient ICA Algorithm with Weighted Orthogonalization

To improve the stability of the traditional natural gradient independent component analysis (ICA) algorithm and the accuracy of its separated results, a adaptive step-size natural gradient ICA algorithm with weighted orthogonalization is proposed. First, to take advantage of the pre-whitening pre-processing and keep the equivariance property of the ICA algorithm, based on the weighted orthogonal constraint on the separating matrix without pre-whitening of observed signals, weighted orthogonalization is introduced after the traditional gradient update. Then, according to the error estimation from the smoothed distance between separated outputs and optimal outputs, we obtain two adaptive step sizes based, respectively, on an unconstrained natural gradient ICA process and a weighted orthogonalization process. Simulation experiment results show that the speed of convergence of the adaptive step-size natural gradient ICA algorithms with weighted orthogonalization are faster than the traditional one; also, the stability of the algorithms and the accuracy of the separated results are improved observably.     

Automated processing of the single-lead electrocardiogram for the detection of obstructive sleep apnoea

A method for the automatic processing of the electrocardiogram (ECG) for the detection of obstructive apnoea is presented. The method screens nighttime single-lead ECG recordings for the presence of major sleep apnoea and provides a minute-by-minute analysis of disordered breathing. A large independently validated database of 70 ECG recordings acquired from normal subjects and subjects with obstructive and mixed sleep apnoea, each of approximately eight hours in duration, was used throughout the study. Thirty-five of these recordings were used for training and 35 retained for independent testing. A wide variety of features based on heartbeat intervals and an ECG-derived respiratory signal were considered. Classifiers based on linear and quadratic discriminants were compared. Feature selection and regularization of classifier parameters were used to optimize classifier performance. Results show that the normal recordings could be separated from the apnoea recordings with a 100% success rate and a minute-by-minute classification accuracy of over 90% is achievable.     

ECG biometric analysis in cardiac irregularity conditions

Biometric traits offer direct solutions to the critical security concerns involved in identity authentication systems. In this paper, a systematic analysis of the electrocardiogram (ECG) signal for application in human recognition is reported, suggesting that cardiac electrical activity is highly personalized in a population. Features extracted from the autocorrelation of healthy ECG signals embed considerable diacritical power, and render fiducial detection unnecessary. The central consideration of this paper is the evaluation of an identification system that is robust to common cardiac irregularities such as premature ventricular contraction (PVC) and atrial premature contraction (APC). Criteria concerning the power distribution and complexity of ECG signals are defined to bring to light abnormal ECG recordings, which are not employable for identification. Experimental results indicate a recognition rate of 96.2% and render identification based on ECG signals rather promising.     

非线性盲源分离的原理及算法综述

本文主要阐述了非线性盲源分离(BSS)/独立成分分析(ICA)模型的基本数学原理、分离算法、算法性能及其应用。首先对线性和非线性BSS/ICA的数学模型作了介绍,重点介绍了非线性BSS/ICA解的不确定性,然后在此基础上对近十年来出现的各种非线性BSS/ICA算法进行简单综述,着重分析了一类可解且应用比较广泛的非线性BSS/ICA模型-后非线性BSS/ICA模型及其分离算法。最后对非线性BSS/ICA存在的问题和发展趋势进行了总结。     

An Enhanced ICA Based on Minimum BER Criterion and Nesterov-Accelerated Adaptive Moment Estimation

Wireless Personal Communications - In this paper, an enhanced independent component analysis (ICA) method is proposed for blind separation of noisy mixture signals. Considering the conventional ICA...