Identification of EEG patterns occurring in anesthesia by means of autoregressive parameters.

In EEG analysis an automatic pattern recognition is of interest. In this paper the usefulness of autoregressive parameters to classify EEG segments recorded during anesthesia is examined. Assuming that the AR parameters are multivariate normally distributed, parametric methods of discriminant analysis can be applied. The results show that AR parameters have high discriminating power and that the lowest error classification rate (smaller than 3%) is obtained by using quadratic discriminant functions. Consequently autoregressive parameters are efficient for classifying EEG segments into general stages of anesthesia.     

The use of a multi-channel kalman filter algorithm in structural analysis of the epileptic EEG

An extension of the Kalman filter algorithm to the multi-channel case is presented and its application as a segmenting procedure in the analysis of the epileptic EEG is discussed. An analytical example of structural analysis, using the segments extracted by the proposed filter, is presented for a particular set of 4-channel EEG recordings. This analysis is shown to be especially fruitful if the autoregressive coefficients - a by product of the filtering procedure - are used to estimate the information flow between the channels by the calculation of partial as well as directed coherences for the representative segments.     

Non-linear and linear forecasting of the EEG time series

The method of non-linear forecasting of time series was applied to different simulated signals and EEG in order to check its ability of distinguishing chaotic from noisy time series. The goodness of prediction was estimated, in terms of the correlation coefficient between forecasted and real time series, for non-linear and autoregressive (AR) methods. For the EEG signal both methods gave similar results. It seems that the EEG signal, in spite of its chaotic character, is well described by the AR model.     

Reduction of the number of recorded EEG channels for routine monitoring in the intensive care unit]

Monitoring patients in the intensive care unit with the aid of the conventional electroencephalogram employing a large number of recording channels is rather difficult, and can be laborious. This imposes limits on the routine application of this method. To investigate the possibility of developing a new monitoring device for easier application in the ICU, we aimed to establish whether the relevant information provided by a multi-channel EEG could be found in a subgroup of channels, thus reducing the number of channels required. Preferably those channels should be identified for use which are least contaminated by artefacts under routine conditions in the ICU. A total of 150 EEG recordings from the intensive care unit were inspected visually for the presence of artefacts. The derivations C3-P3 and C4-P4 proved to be least contaminated, at 35% and 39%, respectively. In these derivations visual assessment of the EEG was found to be impossible due to artefacts in only 4 and 5%, of all cases, respectively. A data set comprising 52 EEG segments with the fewest possible artefacts, was analysed using time series methods. On the basis of multivariate autoregressive processes, a measure was derived which describes the loss of information associated with a reduction in the number of EEG channels. The computation of the information loss for several channel combinations revealed that the derivations F3-C3, C3-P3 and A1-Cz represent a good compromise between information content, number of channels and frequency of artefacts. Practical experience shows that, at least for the control of sedation, a further reduction to a single channel should be possible.(ABSTRACT TRUNCATED AT 250 WORDS)     

Enhancing the dissimilarity-based classification of birdsong recordings

Classification of birdsong recordings can be naturally formulated as a multiple instance problem, where bags of instances are represented by either features or dissimilarities. In bioacoustics, bags typically correspond to regions of interest in spectrograms, which are detected after a segmentation stage of the audio recordings. In this paper, we use different dissimilarity measures between bags and explore whether the subsequent application of metric learning/adaptation methods and the construction of dissimilarity spaces allow increasing the classification performance of birdsong recordings. A publicly available bioacoustic data set is used for the experiments. Our results suggest, in the first place, that appropriate dissimilarity measures are those which capture most of the overall differences between bags, such as the modified Hausdorff distance and the mean minimum distance; in the second place, they confirm the benefit from adapting the applied dissimilarity measure as well as the potential further enhancement of the classification performance by building dissimilarity spaces and increasing training set sizes.     

Correlation dimension based lossless compression of EEG signals

Transmission of long duration EEG signals without loss of information is essential for telemedicine based applications. In this work, a lossless compression scheme for EEG signals based on neural network predictors using the concept of correlation dimension (CD) is proposed. EEG signals which are considered as irregular time series of chaotic processes can be characterized by the non-linear dynamic parameter CD which is a measure of the correlation among the EEG samples. The EEG samples are first divided into segments of 1 s duration and for each segment, the value of CD is calculated. Blocks of EEG samples are then constructed such that each block contains segments with closer CD values. By arranging the EEG samples in this fashion, the accuracy of the predictor is improved as it makes use of highly correlated samples. As a result, the magnitude of the prediction error decreases leading to less number of bits for transmission. Experiments are conducted using EEG signals recorded under different physiological conditions. Different neural network predictors as well as classical predictors are considered. Experimental results show that the proposed CD based preprocessing scheme improves the compression performance of the predictors significantly.     

Comparison of different methods of time shift measurement in EEG

Digital signal processing techniques are often used for measurement of small time shifts between EEG signals. In our work we tested properties of linear cross-correlation and phase/coherence method. The last mentioned method was used in two versions. The first version used fast Fourier transform (FFT) algorithm and the second was based on autoregressive modeling with fixed or adaptive model order. Methods were compared on several testing signals mimicking real EEG signals. The accuracy index for each method was computed. Results showed that for long signal segments all methods bring comparably good results. Accuracy of FFT phase/coherence method significantly decreased when very short segments were used and also decreased with an increasing level of the additive noise. The best results were obtained with autoregressive version of phase/coherence. This method is more reliable and may be used with high accuracy even in very short signals segments and it is also resistant to additive noise.     

基于优化核参数支持向量机的意识任务分类

根据支持向量机的基本原理,给出一种推广误差上界估计判据,并利用该判据进行最优核参数的自动选取。对三种不同意识任务的脑电信号进行多变量自回归模型参数估计,作为意识任务的特征向量,利用支持向量机进行训练和分类测试。分类结果表明,优化核参数的支持向量机分类器取得了最佳的分类效果,分类正确率明显高于径向基函数神经网络。     

Application of recurrence quantification analysis for the automated identification of epileptic EEG signals

Epilepsy is a common neurological disorder that is characterized by the recurrence of seizures. Electroencephalogram (EEG) signals are widely used to diagnose seizures. Because of the non-linear and dynamic nature of the EEG signals, it is difficult to effectively decipher the subtle changes in these signals by visual inspection and by using linear techniques. Therefore, non-linear methods are being researched to analyze the EEG signals. In this work, we use the recorded EEG signals in Recurrence Plots (RP), and extract Recurrence Quantification Analysis (RQA) parameters from the RP in order to classify the EEG signals into normal, ictal, and interictal classes. Recurrence Plot (RP) is a graph that shows all the times at which a state of the dynamical system recurs. Studies have reported significantly different RQA parameters for the three classes. However, more studies are needed to develop classifiers that use these promising features and present good classification accuracy in differentiating the three types of EEG segments. Therefore, in this work, we have used ten RQA parameters to quantify the important features in the EEG signals.These features were fed to seven different classifiers: Support vector machine (SVM), Gaussian Mixture Model (GMM), Fuzzy Sugeno Classifier, K-Nearest Neighbor (KNN), Naive Bayes Classifier (NBC), Decision Tree (DT), and Radial Basis Probabilistic Neural Network (RBPNN). Our results show that the SVM classifier was able to identify the EEG class with an average efficiency of 95.6%, sensitivity and specificity of 98.9% and 97.8%, respectively.     

Application of a dissimilarity index of EEG and its sub-bands on prediction of induced epileptic seizures from rat's EEG signals

ObjectiveEpileptic seizures are defined as manifest of excessive and hyper-synchronous activity of neurons in the cerebral cortex that cause frequent malfunction of the human central nervous system. Therefore, finding precursors and predictors of epileptic seizure is of utmost clinical relevance to reduce the epileptic seizure induced nervous system malfunction consequences. Researchers for this purpose may even guide us to a deep understanding of the seizure generating mechanisms. The goal of this paper is to predict epileptic seizures in epileptic rats.MethodsSeizures were induced in rats using pentylenetetrazole (PTZ) model. EEG signals in interictal, preictal, ictal and postictal periods were then recorded and analyzed to predict epileptic seizures. Epileptic seizures were predicted by calculating an index in consecutive windows of EEG signal and comparing the index with a threshold. In this work, a newly proposed dissimilarity index called Bhattacharyya Based Dissimilarity Index (BBDI), dynamical similarity index and fuzzy similarity index were investigated.ResultsBBDI, dynamical similarity index and fuzzy similarity index were examined on case and control groups and compared to each other. The results show that BBDI outperforms dynamical and fuzzy similarity indices. In order to improve the results, EEG sub-bands were also analyzed. The best result achieved when the proposed dissimilarity index was applied on Delta sub-band that predicts epileptic seizures in all rats with a mean of 299.5 s.ConclusionThe dissimilarity of neural network activity between reference window and present window of EEG signal has a significant increase prior to an epileptic seizure and the proposed dissimilarity index (BBDI) can reveal this variation to predict epileptic seizures. In addition, analyzing EEG sub-bands results in more accurate information about constituent neuronal activities underlying the EEG since certain changes in EEG signal may be amplified when each sub-band is analyzed separately.SignificanceThis paper presents application of a dissimilarity index (BBDI) on EEG signals and its sub-bands to predict PTZ-induced epileptic seizures in rats. Based on the results of this work, BBDI will predict epileptic seizures more accurately and more reliably compared to current indices that increases epileptic patient comfort and improves patient outcomes.     

Simulation of nonstationary EEG

In this paper we present a systematic method for generating simulations of nonstationary EEG. Such simulations are needed, for example, in the evaluation of tracking algorithms. First a state evolution process is simulated. The states are initially represented as segments of stationary autoregressive processes which are described with the corresponding predictor coefficients and prediction error variances. These parameters are then concatenated to give a piecewise time-invariant parameter evolution. The evolution is projected onto an appropriately selected set of smoothly time-varying functions. This projection is used to generate the final EEG simulation. As an example we use this method to simulate the EEG of a drowsy rat. This EEG can be described as toggling between two states that differ in the degree of synchronization of the activity-inducing neuron clusters. Received: 22 June 1994 / Accepted in revised form: 18 February 1997     

A study on the best order for autoregressive EEG modelling

The autoregressive (AR) model is a widely used tool in electroencephalogram (EEG) analysis. The dependence of the AR model on both the segment length and several characteristic EEG patterns is addressed. The best AR model order is computed with three different criteria. The results show that the Rissanen criteria provides the more consistent order estimate for the EEG patterns considered. This study shows that for our data set, a 5th order AR model represents adequately 1- or 2-s EEG segments with the exception of featureless background, where higher order models are necessary.     

An autoregressive method for the measurement of synchronization of interictal and ictal EEG signals

We propose a new measure of synchronization of multichannel ictal and interictal EEG signals. The measure is based on the residual covariance matrix of a multichannel autoregressive model. A major advantage of this measure is its ability to be interpreted both in the framework of stochastic and deterministic models. A preliminary analysis of EEG data from three patients using this measure documents the expected increased synchronization during ictal periods but also reveals that increased synchrony persists for prolonged periods (up to 2 h or more) in the postictal period. Received: 20 July 1997 / Accepted in revised form: 26 January 1999     

Dissimilarity for partially ranked data and its application to cover-abundance data

Cover-abundance estimates are commonly employed in phytosociological investigations to record the performance of species. Because the coded values are on an ordinal scale of measure, various authors have suggested that some transformation is necessary before such values can be used for classification and ordination. However, it is not clear that transformation is a sufficient treatment, and it would seem preferable to use ordinal data directly. In this paper we examine such direct use of partial rankings and show that several dissimilarity measures can be defined for this case without invoking any transformations. They include dissimilarity measures associated with various rank correlation measures and with distances between strings; all the measure are variant forms of Hausdorf's interset distance. Certain other kinds of data, such as those employing dominant and subdominant species and the dry-weight-rank estimation of biomass, are also on an ordinal scale and could be analysed using similar techniques.To illustrate the approach, a string dissimilarity measure is used to analyse a set of data from Slovakian grasslands which appear to reflect a simple gradient. The original data were recorded with 10 classes of performance and are analysed using hierarchical and nondeterministic, overlapping, classifications.     

The adaptive classification of the dynamic spectral patterns in the human EEG]

The classification technique of single spectra based on a matrix of intercorrelation between these spectra and the fixed set of standard spectral patterns (SP) has been put forward. Including in the classification technique a special procedure for automatic adaptation of the standard SP to given EEG records makes it possible to reduce the number of unclassified single spectra to a minimum (6-10%), which we can ignore during comparative analysis of the EEG classification profiles. Using the universal set of standard SP makes it possible to compare the results of classification of different EEG records. The results of the analysis of classification profiles of human multichannel EEG during performance of the memory task on perception of visual images are described in the paper. It has been shown that both the total EEG reorganization associated with the alpha rhythm blockade during eyes opening and less noticeable EEG shifts accompanying changes in the stages of cognitive activity are underlain by a rather differentiated transformations of relative contributions of each type of the SP into the total power spectrum. It has been revealed that a relatively small part (15-20%) of the elementary EEG segments participates in the reorganization of the EEG classification profile.     

Power spectral density and coherence analysis of Alzheimer’s EEG

In this paper, we investigate the abnormalities of electroencephalograph (EEG) signals in the Alzheimer’s disease (AD) by analyzing 16-scalp electrodes EEG signals and make a comparison with the normal controls. The power spectral density (PSD) which represents the power distribution of EEG series in the frequency domain is used to evaluate the abnormalities of AD brain. Spectrum analysis based on autoregressive Burg method shows that the relative PSD of AD group is increased in the theta frequency band while significantly reduced in the alpha2 frequency bands, particularly in parietal, temporal, and occipital areas. Furthermore, the coherence of two EEG series among different electrodes is analyzed in the alpha2 frequency band. It is demonstrated that the pair-wise coherence between different brain areas in AD group are remarkably decreased. Interestingly, this decrease of pair-wise electrodes is much more significant in inter-hemispheric areas than that in intra-hemispheric areas. Moreover, the linear cortico-cortical functional connectivity can be extracted based on coherence matrix, from which it is shown that the functional connections are obviously decreased, the same variation trend as relative PSD. In addition, we combine both features of the relative PSD and the normalized degree of functional network to discriminate AD patients from the normal controls by applying a support vector machine model in the alpha2 frequency band. It is indicated that the two groups can be clearly classified by the combined feature. Importantly, the accuracy of the classification is higher than that of any one feature. The obtained results show that analysis of PSD and coherence-based functional network can be taken as a potential comprehensive measure to distinguish AD patients from the normal, which may benefit our understanding of the disease.     

Automatic classification of EEG segments and extraction of representative ones by dynamic clusters method

Use of the dynamic clusters method for automatic extraction of compressed information about recorded EEG signal is presented. The computer first divides the record into quasi-stationary segments by means of adaptive segmentation. Second, the extracted segments are classified by a method of dynamic clusters into homogeneous classes. One part of the used clustering algorithm permits to specify and draw the most typical class members, which may represent the whole studied EEG signal and may be used as input for the further phase of the automatic EEG analysis, i.e. for the classification of the whole EEG records. The above procedure was applied to a 75 sec long EEG record of anaesthetized cat intoxicated by CO.     

Multiple characteristics analysis of Alzheimer’s electroencephalogram by power spectral density and Lempel–Ziv complexity

To investigate the electroencephalograph (EEG) background activity in patients with Alzheimer’s disease (AD), power spectrum density (PSD) and Lempel–Ziv (LZ) complexity analysis are proposed to extract multiple effective features of EEG signals from AD patients and further applied to distinguish AD patients from the normal controls. Spectral analysis based on autoregressive Burg method is first used to quantify the power distribution of EEG series in the frequency domain. Compared with the control group, the relative PSD of AD group is significantly higher in the theta frequency band while lower in the alpha frequency bands. In order to explore the nonlinear information, Lempel–Ziv complexity (LZC) and multi-scale LZC is further applied to all electrodes for the four frequency bands. Analysis results demonstrate that the group difference is significant in the alpha frequency band by LZC and multi-scale LZC analysis. However, the group difference of multi-scale LZC is much more remarkable, manifesting as more channels undergo notable changes, particularly in electrodes O1 and O2 in the occipital area. Moreover, the multi-scale LZC value provided a better classification between the two groups with an accuracy of 85.7 %. In addition, we combine both features of the relative PSD and multi-scale LZC to discriminate AD patients from the normal controls by applying a support vector machine model in the alpha frequency band. It is indicated that the two groups can be clearly classified by the combined feature. Importantly, the accuracy of the classification is higher than that of any one feature, reaching 91.4 %. The obtained results show that analysis of PSD and multi-scale LZC can be taken as a potential comprehensive measure to distinguish AD patients from the normal controls, which may benefit our understanding of the disease.     

Modeling of movement-related potentials using a fractal approach

In bio-signal applications, classification performance depends greatly on feature extraction, which is also the case for electroencephalogram (EEG) based applications. Feature extraction, and consequently classification of EEG signals is not an easy task due to their inherent low signal-to-noise ratios and artifacts. EEG signals can be treated as the output of a non-linear dynamical (chaotic) system in the human brain and therefore they can be modeled by their dimension values. In this study, the variance fractal dimension technique is suggested for the modeling of movement-related potentials (MRPs). Experimental data sets consist of EEG signals recorded during the movements of right foot up, lip pursing and a simultaneous execution of these two tasks. The experimental results and performance tests show that the proposed modeling method can efficiently be applied to MRPs especially in the binary approached brain computer interface applications aiming to assist severely disabled people such as amyotrophic lateral sclerosis patients in communication and/or controlling devices.     

Two-stage biometric authentication method using thought activity brain waves

Brain waves are proposed as a biometric for verification of the identities of individuals in a small group. The approach is based on a novel two-stage biometric authentication method that minimizes both false accept error (FAE) and false reject error (FRE). These brain waves (or electroencephalogram (EEG) signals) are recorded while the user performs either one or several thought activities. As different individuals have different thought processes, this idea would be appropriate for individual authentication. In this study, autoregressive coefficients, channel spectral powers, inter-hemispheric channel spectral power differences, inter-hemispheric channel linear complexity and non-linear complexity (approximate entropy) values were used as EEG features by the two-stage authentication method with a modified four fold cross validation procedure. The results indicated that perfect accuracy was obtained, i.e. the FRE and FAE were both zero when the proposed method was tested on five subjects using certain thought activities. This initial study has shown that the combination of the two-stage authentication method with EEG features from thought activities has good potential as a biometric as it is highly resistant to fraud. However, this is only a pilot type of study and further extensive research with more subjects would be necessary to establish the suitability of the proposed method for biometric applications.