Epileptic Electroencephalogram Research Articles

Epileptic electroencephalogram recognition based on discrete S transform and permutation entropy

Electroencephalogram(EEG) analysis has important reference value in the diagnosis of epilepsy. The automatic classification of epileptic EEG can be used to judge the patient's situation in time,which is of great significance in clinical application. In order to solve the problem that the recognition accuracy is not high by using the single feature of EEG signals and avoid the influence of wavelet basis function selection on recognition results,a method of automatic discrimination of epileptic EEG signals based on S transform and permutation entropy is proposed. Firstly, the original signals are decomposed by discrete S transform, and then we calculate the fluctuation index of coefficients of each rhythm and combine the permutation entropy of EEG signals into a feature vector and use Real AdaBoost classifier to discriminate the epileptic EEG signals in muti-period. In this study, we used the epilepsy database from University of Bonn. Three groups of EEG signals, including the data of normal people with their eyes open, the data collected inside of the epileptic foci from patients during their interictal period and the data during their ictal period, were used to test effectiveness. The results of this study showed that the fluctuation index of each rhythm could be used to characterize the normal, interictal and ictal epileptic EEG signals effectively, and the recognition accuracy of multiple features was much higher than that of single feature. The average recognition accuracy could reach 98.13%. Compared with time-frequency feature extraction method or nonlinear feature extraction method only,the recognition accuracy was increased by more than 1.2% and 8.1% respectively, which was superior to the methods recorded in many other literatures. Therefore, this method has a good application prospect in diagnosis of epilepsy.

Reconstructing multi-mode networks from multivariate time series

Unveiling the dynamics hidden in multivariate time series is a task of the utmost importance in a broad variety of areas in physics. We here propose a method that leads to the construction of a novel functional network, a multi-mode weighted graph combined with an empirical mode decomposition, and to the realization of multi-information fusion of multivariate time series. The method is illustrated in a couple of successful applications (a multi-phase flow and an epileptic electro-encephalogram), which demonstrate its powerfulness in revealing the dynamical behaviors underlying the transitions of different flow patterns, and enabling to differentiate brain states of seizure and non-seizure.

Chaos feature study in fractional Fourier domain for preictal prediction of epileptic seizure

Epileptic seizure prediction is limited by the unstable performance of suboptimal models. Studies of new methods for reliable preictal prediction have significant impact on the control, early care and online treatment of epileptic seizure. The traditional chaos measure does not effectively identify multiple states of epileptic electroencephalogram (EEG). A novel method was adopted to capture subtle chaotic dynamics for epileptic signals in fractional Fourier transform domain. Algorithm of the largest Lyapunov exponent was modified to adapt the transformed series by using an energy measure to determine appropriate fractional order. The performance of our proposed method was evaluated with an automatic model of preictal prediction using artificial neural networks as classifier. The results showed that the new model yielded higher accuracy in identifying the preictal state compared to the original largest Lyapunov exponent. Experimental results with noisy scalp epileptic EEGs also demonstrated the potential and robustness of our approach to discriminate preictal from interictal and ictal states, and it provided a novel methodology for reliable preictal prediction of epileptic seizure.

Recognition of Epileptic EEG Signals Using a Novel Multiview TSK Fuzzy System

Recognition of epileptic electroencephalogram (EEG) signals using machine learning techniques is becoming popular. In general, the construction of intelligent epileptic EEG recognition system involves two steps. First, an appropriate feature extraction method is applied to obtain representative features from the original raw EEG signals. Second, an effective intelligent model is trained based on the extracted features. However, there exist two major challenges in the process: 1) it is nontrivial to determine the appropriate feature extraction method to be used; 2) although many classical machine learning methods have been used for epileptic EEG recognition, most of them are “black box” approaches and more interpretable methods are desirable. To address these two challenges, a new epileptic EEG recognition method based on a multiview learning framework and fuzzy system modeling is proposed. First, multiview EEG data are generated by employing different feature extraction methods to obtain the features from different views of the signals. Second, the classical Takagi-Sugeno-Kang fuzzy system (TSK-FS) is introduced as an easy-to-interpret recognition model to develop a multiview TSK-FS method, called MV-TSK-FS, to identify epileptic EEG signals. For the proposed MV-TSK-FS, the importance of each view, i.e., the importance of each feature extraction method, can be evaluated according to the weighting of each view, and consequently the final decision can be made based on the weighted outputs of different views. Experimental results indicate that the MV-TSK-FS is a promising method when compared with the state-of-the-art algorithms.

Statistical features-based comparison of analysis and synthesis of normal and epileptic electroencephalograms for various wavelets

Statistical features-based comparison of analysis and synthesis of normal and epileptic electroencephalograms for various wavelets

Enhancing perceptrons with contrastive biclusters

An enhanced perceptron model, referred to as BicNeuron (BN), which is based on the novel concept of contrastive biclusters is investigated. According to this viewpoint, a coherent bicluster (i.e. a subset of data patterns showing high similarity across a particular subset of features) belonging to one class is required to show high discrimination from the nearest patterns of the other class when projected on the same feature set. On each local subspace associated with a different pair of contrastive biclusters a perceptron is induced and the model with highest area under the receiver operating characteristic curve value is selected as the final classifier. Experiments involving the discrimination of epileptic and non-epileptic electroencephalogram (EEG) signals show that the BN approach can be very useful, prevailing significantly on standard and kernel perceptrons in terms of classification accuracy.

AR based quadratic feature extraction in the VMD domain for the automated seizure detection of EEG using random forest classifier

Abstract Visual inspection of epileptic electroencephalogram (EEG) by neurologists is time-consuming and tedious. To overcome the problems, numerous automated seizure detection techniques, combining signal processing and machine learning, have been developed. Although 100% accuracy has been achieved for classifying non-seizure and seizure EEG records in up-to-date articles, the result of distinguishing normal, interictal and ictal EEG is still not satisfying. In this paper, a fusion method of variational mode decomposition (VMD) and autoregression (AR) based quadratic feature extraction was proposed for feature extraction and the random forest classifier was employed to hand with three-classification task. The raw EEG was decomposed into a fixed number of band-limited intrinsic mode functions (BLIMFs) using VMD, then a logarithmic operation was imposed on each BLIMF. Subsequently, optimal AR based quadratic feature extraction was conducted on all the BLIMFs and the extracted feature vectors were fed into random forest classifier for classification. Experimental results on the Bonn epilepsy EEG dataset show that the best accuracy of the proposed scheme is 97.352% and it outperforms than the fixed-order AR based feature extraction technique. The developed technology is proven efficient for seizure detection. It can be further programmed into software and the software can be applied in hospitals to assist the neurologists for seizure detection.

Takagi–Sugeno–Kang Transfer Learning Fuzzy Logic System for the Adaptive Recognition of Epileptic Electroencephalogram Signals

The intelligent recognition of electroencephalogram (EEG) signals has become an important approach to the detection of epilepsy. Among existing intelligent identification methods, fuzzy logic systems (FLSs) have shown a distinctive advantage in identifying epileptic EEG signals because of their strong learning abilities and interpretability. Like many conventional intelligent methods for recognizing EEG signals, in the training of FLS, it is assumed that the training dataset and test dataset are drawn from data that are identically distributed. However, this assumption is not necessarily valid in practice as it is not uncommon for the two datasets to have different distributions. To overcome this problem, a strategy is presented in this paper to construct a Takagi–Sugeno–Kang (TSK) FLS based on transductive transfer learning for identifying epileptic EEG signals. Two novel objective functions, achieved by integrating the transductive transfer learning mechanism, are proposed for the training of the TSK FLS. As regression and binary classification are two common approaches to multiclass classification, the TSK transfer learning FLS algorithms for regression and binary classification are developed, respectively, to construct the corresponding TSK FLS. Both algorithms are further used to perform a multiclass classification to recognize epileptic EEG signals. Their performance in the epileptic EEG datasets indicates promise in dealing with situations where the training and test datasets differ with regard to data distribution.

The effects of carbamazepine in the intrahippocampal kainate model of temporal lobe epilepsy depend on seizure definition and mouse strain.

SummaryObjectiveMesial temporal lobe epilepsy (TLE) with hippocampal sclerosis is a predominant form of acquired epilepsy, characterized by recurrent simple and complex partial seizures that are often resistant to treatment. Mice developing spontaneous recurrent nonconvulsive and convulsive seizures after intrahippocampal injection of the excitotoxic glutamate agonist kainate are thought to represent a valuable model of mesial TLE. Epileptic electroencephalogram (EEG) activity recorded in this model from the kainate focus in the ipsilateral hippocampus is resistant to antiseizure drugs such as carbamazepine (CBZ). We compared the efficacy of CBZ in this model in two different mouse strains (FVB/N and NMRI). Furthermore, we evaluated whether changes in the definition of electrographic seizures affect the antiseizure efficacy of CBZ.MethodsAs in previous studies, two types of epileptic EEG activity were defined: high‐voltage sharp waves (HVSWs) and hippocampal paroxysmal discharges (HPDs). The characteristics of these paroxysmal EEG events in epileptic mice were compared with EEG criteria for nonconvulsive seizures in patients. For HVSWs, different spike frequencies, interevent intervals, and amplitudes were used as inclusion and exclusion criteria. In addition to CBZ, some experiments were performed with diazepam (DZP) and phenobarbital (PB).ResultsFemale epileptic FVB/N mice predominantly exhibited frequent HVSWs, but only infrequent HPDs or secondarily generalized convulsive seizures. Slight changes in HVSW definition determined whether they were resistant or responsive to CBZ. Male NMRI mice exhibited both HVSWs and HPDs. HVSWs were more resistant than HPDs to suppression by CBZ. Both types of epileptic EEG activity were rapidly suppressed by DZP and PB.SignificanceThe data demonstrate that focal electrographic seizures in the intrahippocampal kainate mouse model are less resistant than previously thought. Both mouse strain and the criteria chosen for definition of EEG seizures determine whether such seizures are drug‐resistant or ‐responsive.

Increment Entropy as a Measure of Complexity for Time Series

Entropy has been a common index to quantify the complexity of time series in a variety of fields. Here, we introduce an increment entropy to measure the complexity of time series in which each increment is mapped onto a word of two letters, one corresponding to the sign and the other corresponding to the magnitude. Increment entropy (IncrEn) is defined as the Shannon entropy of the words. Simulations on synthetic data and tests on epileptic electroencephalogram (EEG) signals demonstrate its ability of detecting abrupt changes, regardless of the energetic (e.g., spikes or bursts) or structural changes. The computation of IncrEn does not make any assumption on time series, and it can be applicable to arbitrary real-world data.

Intermittency of intermittencies in characteristic oscillatory patterns on epileptic electroencephalograms

The temporal dynamics of characteristic oscillatory patterns in an epileptic EEG are analyzed. A statistical analysis of time gaps between oscillatory events on EEG records is performed following the automatic segmentation of EEG signals. Both sleep spindles and spike-wave discharges exhibit intermittency dynamics, and the collective dynamics of these patterns is characterized by the intermittency of intermittencies.

Epileptic EEG visualization and sonification based on linear discriminate analysis.

In this paper, we first presents a high accuracy epileptic electroencephalogram (EEG) classification algorithm. EEG data of epilepsy patients are preprocessed, segmented, and decomposed to intrinsic mode functions, from which features are extracted. Two classifiers are trained based on linear discriminant analysis (LDA) to classify EEG data into three types, i.e., normal, spike, and seizure. We further in-depth investigate the changes of the decision values in LDA on continuous EEG data. An epileptic EEG visualization and sonification algorithm is proposed to provide both temporal and spatial information of spike and seizure of epilepsy patients. In the experiment, EEG data of six subjects (two normal and four seizure patients) are included. The experiment result shows the proposed epileptic EEG classification algorithm achieves high accuracy. As well, the visualization and sonification algorithm exhibits a great help in nursing seizure patients and localizing the area of seizures.

The detection of epileptic seizure signals based on fuzzy entropy

BackgroundEntropy is a nonlinear index that can reflect the degree of chaos within a system. It is often used to analyze epileptic electroencephalograms (EEG) to detect whether there is an epileptic attack. Much research into the state inspection of epileptic seizures has been conducted based on sample entropy (SampEn). However, the study of epileptic seizures based on fuzzy entropy (FuzzyEn) has lagged behind. New methodsWe propose a method of state inspection of epileptic seizures based on FuzzyEn. The method first calculates the FuzzyEn of EEG signals from different epileptic states, and then feature selection is conducted to obtain classification features. Finally, we use the acquired classification features and a grid optimization method to train support vector machines (SVM). ResultsThe results of two open-EEG datasets in epileptics show that there are major differences between seizure attacks and non-seizure attacks, such that FuzzyEn can be used to detect epilepsy, and our method obtains better classification performance (accuracy, sensitivity and specificity of classification of the CHB-MIT are 98.31%, 98.27% and 98.36%, and of the Bonn are 100%, 100%, 100%, respectively). Comparisons with existing method(s)To verify the performance of the proposed method, a comparison of the classification performance for epileptic seizures using FuzzyEn and SampEn is conducted. Our method obtains better classification performance, which is superior to the SampEn-based methods currently in use. ConclusionsThe results indicate that FuzzyEn is a better index for detecting epileptic seizures effectively. The FuzzyEn-based method is preferable, exhibiting potential desirable applications for medical treatment.

Examination of the wavelet-based approach for measuring self-similarity of epileptic electroencephalogram data

Self-similarity or scale-invariance is a fascinating characteristic found in various signals including electroencephalogram (EEG) signals. A common measure used for characterizing self-similarity or scale-invariance is the spectral exponent. In this study, a computational method for estimating the spectral exponent based on wavelet transform was examined. A series of Daubechies wavelet bases with various numbers of vanishing moments were applied to analyze the self-similar characteristics of intracranial EEG data corresponding to different pathological states of the brain, i.e., ictal and interictal states, in patients with epilepsy. The computational results show that the spectral exponents of intracranial EEG signals obtained during epileptic seizure activity tend to be higher than those obtained during non-seizure periods. This suggests that the intracranial EEG signals obtained during epileptic seizure activity tend to be more self-similar than those obtained during non-seizure periods. The computational results obtained using the wavelet-based approach were validated by comparison with results obtained using the power spectrum method.

Transductive domain adaptive learning for epileptic electroencephalogram recognition.

Intelligent recognition of electroencephalogram (EEG) signals is an important means for epilepsy detection. Almost all conventional intelligent recognition methods assume that the training and testing data of EEG signals have identical distribution. However, this assumption may indeed be invalid for practical applications due to differences in distributions between the training and testing data, making the conventional epilepsy detection algorithms not feasible under such situations. In order to overcome this problem, we proposed a transfer-learning-based intelligent recognition method for epilepsy detection. We used the large-margin-projected transductive support vector machine method (LMPROJ) to learn the useful knowledge between the training domain and testing domain by calculating the maximal mean discrepancy. The method can effectively learn a model for the testing data with training data of different distributions, thereby relaxing the constraint that the data distribution in the training and testing samples should be identical. The experimental validation is performed over six datasets of electroencephalogram signals with three feature extraction methods. The proposed LMPROJ-based transfer learning method was compared with five conventional classification methods. For the datasets with identical distribution, the performance of these six classification methods was comparable. They all could achieve an accuracy of 90%. However, the LMPROJ method obviously outperformed the five conventional methods for experimental datasets with different distribution between the training and test data. Regardless of the feature extraction method applied, the mean classification accuracy of the proposed method is above 93%, which is greater than that of the other five methods with statistical significance. The proposed transfer-learning-based method has better classification accuracy and adaptability than the conventional methods in classifying EEG signals for epilepsy detection.

Examination of scale-invariant characteristics of epileptic electroencephalograms using wavelet-based analysis

There is evidence that biological and physiological systems including the brain exhibit can exhibit fractal characteristics that can be used to identify the state of the system. In this study, wavelet-based fractal analysis is used to examine self-similar or scale-invariant characteristics of intracranial EEG data in terms of the spectral exponent. The intracranial EEG data were recorded from subjects with epilepsy during non-seizure period and during epileptic seizure activity. From the computational results, it is observed that the self-similar or scale-invariant characteristics of the intracranial EEG data obtained during these two periods are significantly different. The actual value of the estimated spectral exponent depends on the wavelet bases used for the computations.

Multiscale permutation entropy analysis of electroencephalogram

We carried out a detailed analysis and a comparison between normal and epileptic electroencephalogram (EEG) based on multiscale permutation entropy. The relationship between multiscale permutation entropy values of EEG and age, and the effect of scale factor on multiscale permutation entropy value were also discussed. By analyzing normal and epileptic EEG based on multiscale permutation entropy, we found that, at the same age, multiscale permutation entropy value of the normal group’s EEG is higher than that of the epileptic group by an average of 0.19, about 7.9%. In addition, for people of age 3 to 35, their multiscale permutation entropies are clearly maximum. When scale factor is smaller than 15, the value of their entropy would reduce no matter whether the age increases or decreases. The results indicate that multiscale permutation entropy can distinguish between normal and epileptic EEG and reflect the general process of human brain development.

Optimal Classification of Epileptic EEG Signals Using Neural Networks and Harmony Search Methods

In this paper, the Harmony Search (HS)-aided BP neural networks are used for the classification of the epileptic electroencephalogram (EEG) signals. It is well known that the gradient descent-based learning method can result in local optima in the training of BP neural networks, which may significantly affect their approximation performances. Three HS methods, the original version and two new variations recently proposed by the authors of the present paper, are applied here to optimize the weights in the BP neural networks for the classification of the epileptic EEG signals. Simulations have demonstrated that the classification accuracy of the BP neural networks can be remarkably improved by the HS method-based training.

Model-Based Spike Detection of Epileptic EEG Data

Accurate automatic spike detection is highly beneficial to clinical assessment of epileptic electroencephalogram (EEG) data. In this paper, a new two-stage approach is proposed for epileptic spike detection. First, the k-point nonlinear energy operator (k-NEO) is adopted to detect all possible spike candidates, then a newly proposed spike model with slow wave features is applied to these candidates for spike classification. Experimental results show that the proposed system, using the AdaBoost classifier, outperforms the conventional method in both two- and three-class EEG pattern classification problems. The proposed system not only achieves better accuracy for spike detection, but also provides new ability to differentiate between spikes and spikes with slow waves. Though spikes with slow waves occur frequently in epileptic EEGs, they are not used in conventional spike detection. Identifying spikes with slow waves allows the proposed system to have better capability for assisting clinical neurologists in routine EEG examinations and epileptic diagnosis.

Review: A Survey of Performance and Techniques for Automatic Epilepsy Detection

Fil: Orosco, Lorena Liliana. Universidad Nacional de San Juan. Facultad de Ingenieria. Departamento de Electronica y Automatica. Gabinete de Tecnologia Medica; Argentina. Consejo Nacional de Investigaciones Cientificas y Tecnicas; Argentina