Epileptic Seizure Detection in EEGs Using Time–Frequency Analysis

Automatic identification of epileptic seizures from EEG signals using linear programming boosting

BACKGROUND AND OBJECTIVE

Computerized epileptic seizure detection is essential for expediting epilepsy diagnosis and research and for assisting medical professionals. Moreover, the implementation of an epilepsy monitoring device that has low power and is portable requires a reliable and successful seizure detection scheme. In this work, the problem of automated epilepsy seizure detection using singe-channel EEG signals has been addressed.

METHODS

At first, segments of EEG signals are decomposed using a newly proposed signal processing scheme, namely complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN). Six spectral moments are extracted from the CEEMDAN mode functions and train and test matrices are formed afterward. These matrices are fed into the classifier to identify epileptic seizures from EEG signal segments. In this work, we implement an ensemble learning based machine learning algorithm, namely linear programming boosting (LPBoost) to perform classification.

RESULTS

The efficacy of spectral features in the CEEMDAN domain is validated by graphical and statistical analyses. The performance of CEEMDAN is compared to those of its predecessors to further inspect its suitability. The effectiveness and the appropriateness of LPBoost are demonstrated as opposed to the commonly used classification models. Resubstitution and 10 fold cross-validation error analyses confirm the superior algorithm performance of the proposed scheme. The algorithmic performance of our epilepsy seizure identification scheme is also evaluated against state-of-the-art works in the literature. Experimental outcomes manifest that the proposed seizure detection scheme performs better than the existing works in terms of accuracy, sensitivity, specificity, and Cohen's Kappa coefficient.

CONCLUSION

It can be anticipated that owing to its use of only one channel of EEG signal, the proposed method will be suitable for device implementation, eliminate the onus of clinicians for analyzing a large bulk of data manually, and expedite epilepsy diagnosis.

LMD Based Features for the Automatic Seizure Detection of EEG Signals Using SVM

Achieving the goal of detecting seizure activity automatically using electroencephalogram (EEG) signals is of great importance and significance for the treatment of epileptic seizures. To realize this aim, a newly-developed time-frequency analytical algorithm, namely local mean decomposition (LMD), is employed in the presented study. LMD is able to decompose an arbitrary signal into a series of product functions (PFs). Primarily, the raw EEG signal is decomposed into several PFs, and then the temporal statistical and non-linear features of the first five PFs are calculated. The features of each PF are fed into five classifiers, including back propagation neural network (BPNN), K-nearest neighbor (KNN), linear discriminant analysis (LDA), un-optimized support vector machine (SVM) and SVM optimized by genetic algorithm (GA-SVM), for five classification cases, respectively. Confluent features of all PFs and raw EEG are further passed into the high-performance GA-SVM for the same classification tasks. Experimental results on the international public Bonn epilepsy EEG dataset show that the average classification accuracy of the presented approach are equal to or higher than 98.10% in all the five cases, and this indicates the effectiveness of the proposed approach for automated seizure detection.

Detection of epileptic seizure in EEG signals using linear least squares preprocessing

An epileptic seizure is a transient event of abnormal excessive neuronal discharge in the brain. This unwanted event can be obstructed by detection of electrical changes in the brain that happen before the seizure takes place. The automatic detection of seizures is necessary since the visual screening of EEG recordings is a time consuming task and requires experts to improve the diagnosis. Much of the prior research in detection of seizures has been developed based on artificial neural network, genetic programming, and wavelet transforms. Although the highest achieved accuracy for classification is 100%, there are drawbacks, such as the existence of unbalanced datasets and the lack of investigations in performances consistency. To address these, four linear least squares-based preprocessing models are proposed to extract key features of an EEG signal in order to detect seizures. The first two models are newly developed. The original signal (EEG) is approximated by a sinusoidal curve. Its amplitude is formed by a polynomial function and compared with the predeveloped spline function. Different statistical measures, namely classification accuracy, true positive and negative rates, false positive and negative rates and precision, are utilised to assess the performance of the proposed models. These metrics are derived from confusion matrices obtained from classifiers. Different classifiers are used over the original dataset and the set of extracted features. The proposed models significantly reduce the dimension of the classification problem and the computational time while the classification accuracy is improved in most cases. The first and third models are promising feature extraction methods with the classification accuracy of 100%. Logistic, LazyIB1, LazyIB5, and J48 are the best classifiers. Their true positive and negative rates are 1 while false positive and negative rates are 0 and the corresponding precision values are 1. Numerical results suggest that these models are robust and efficient for detecting epileptic seizure.

Classification of 5-S Epileptic EEG Recordings Using Distribution Entropy and Sample Entropy

Epilepsy is an electrophysiological disorder of the brain, the hallmark of which is recurrent and unprovoked seizures. Electroencephalogram (EEG) measures electrical activity of the brain that is commonly applied as a non-invasive technique for seizure detection. Although a vast number of publications have been published on intelligent algorithms to classify interictal and ictal EEG, it remains an open question whether they can be detected using short-length EEG recordings. In this study, we proposed three protocols to select 5 s EEG segment for classifying interictal and ictal EEG from normal. We used the publicly-accessible Bonn database, which consists of normal, interical, and ictal EEG signals with a length of 4097 sampling points (23.6 s) per record. In this study, we selected three segments of 868 points (5 s) length from each recordings and evaluated results for each of them separately. The well-studied irregularity measure—sample entropy (SampEn)—and a more recently proposed complexity measure—distribution entropy (DistEn)—were used as classification features. A total of 20 combinations of input parameters m and τ for the calculation of SampEn and DistEn were selected for compatibility. Results showed that SampEn was undefined for half of the used combinations of input parameters and indicated a large intra-class variance. Moreover, DistEn performed robustly for short-length EEG data indicating relative independence from input parameters and small intra-class fluctuations. In addition, it showed acceptable performance for all three classification problems (interictal EEG from normal, ictal EEG from normal, and ictal EEG from interictal) compared to SampEn, which showed better results only for distinguishing normal EEG from interictal and ictal. Both SampEn and DistEn showed good reproducibility and consistency, as evidenced by the independence of results on analysing protocol.

A novel genetic programming approach for epileptic seizure detection

The human brain is a delicate mix of neurons (brain cells), electrical impulses and chemicals, known as neurotransmitters. Any damage has the potential to disrupt the workings of the brain and cause seizures. These epileptic seizures are the manifestations of epilepsy. The electroencephalograph (EEG) signals register average neuronal activity from the cerebral cortex and label changes in activity over large areas. A detailed analysis of these electroencephalograph (EEG) signals provides valuable insights into the mechanisms instigating epileptic disorders. Moreover, the detection of interictal spikes and epileptic seizures in an EEG signal plays an important role in the diagnosis of epilepsy. Automatic seizure detection methods are required, as these epileptic seizures are volatile and unpredictable. This paper deals with an automated detection of epileptic seizures in EEG signals using empirical mode decomposition (EMD) for feature extraction and proposes a novel genetic programming (GP) approach for classifying the EEG signals. Improvements in the standard GP approach are made using a Constructive Genetic Programming (CGP) in which constructive crossover and constructive subtree mutation operators are introduced. A hill climbing search is integrated in crossover and mutation operators to remove the destructive nature of these operators. A new concept of selecting the Globally Prime offspring is also presented to select the best fitness offspring generated during crossover. To decrease the time complexity of GP, a new dynamic fitness value computation (DFVC) is employed to increase the computational speed. We conducted five different sets of experiments to evaluate the performance of the proposed model in the classification of different mixtures of normal, interictal and ictal signals, and the accuracies achieved are outstandingly high. The experimental results are compared with the existing methods on same datasets, and these results affirm the potential use of our method for accurately detecting epileptic seizures in an EEG signal.

Differential diagnosis of squamous cell carcinoma in situ using skin histopathological images

Differential diagnosis of squamous cell carcinoma in situ is of great importance for prognosis and decision making in the disease treatment procedure. Currently, differential diagnosis is done by pathologists based on examination of the histopathological slides under the microscope, which is time consuming and prone to inter and intra observer variability. In this paper, we have proposed an automated method for differential diagnosis of SCC in situ from actinic keratosis, which is known to be a precursor of squamous cell carcinoma. The process begins with epidermis segmentation and cornified layer removal. Then, epidermis axis is specified using the paths in its skeleton and the granular layer is removed via connected components analysis. Finally, diagnosis is done based on the classification result of intensity profiles extracted from lines perpendicular to the epidermis axis. The results of the study are in agreement with the gold standards provided by expert pathologists.

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.

Seizure detection approach using S-transform and singular value decomposition

Automatic seizure detection plays a significant role in the diagnosis of epilepsy. This paper presents a novel method based on S-transform and singular value decomposition (SVD) for seizure detection. Primarily, S-transform is performed on EEG signals, and the obtained time-frequency matrix is divided into submatrices. Then, the singular values of each submatrix are extracted using singular value decomposition (SVD). Effective features are constructed by adding the largest singular values in the same frequency band together and fed into Bayesian linear discriminant analysis (BLDA) classifier for decision. Finally, postprocessing is applied to obtain higher sensitivity and lower false detection rate. A total of 183.07 hours of intracranial EEG recordings containing 82 seizure events from 20 patients were used to evaluate the system. The proposed method had a sensitivity of 96.40% and a specificity of 99.01%, with a false detection rate of 0.16/h.

Epilepsy analytic system with cloud computing

Biomedical data analytic system has played an important role in doing the clinical diagnosis for several decades. Today, it is an emerging research area of analyzing these big data to make decision support for physicians. This paper presents a parallelized web-based tool with cloud computing service architecture to analyze the epilepsy. There are many modern analytic functions which are wavelet transform, genetic algorithm (GA), and support vector machine (SVM) cascaded in the system. To demonstrate the effectiveness of the system, it has been verified by two kinds of electroencephalography (EEG) data, which are short term EEG and long term EEG. The results reveal that our approach achieves the total classification accuracy higher than 90%. In addition, the entire training time accelerate about 4.66 times and prediction time is also meet requirements in real time.

Epileptic Seizure Detection Based on Partial Directed Coherence Analysis

Long-term video EEG epilepsy monitoring can help doctors diagnose and cure epilepsy. The workload of doctors to read the EEG signals of epilepsy patients can be effectively reduced by automatic seizure detection. The application of partial directed coherence (PDC) analysis as mechanism for feature extraction in the scalp EEG recordings for seizure detection could reflect the physiological changes of brain activity before and after seizure onsets. In this study, a new approach on the basis of PDC was proposed to detect the seizure intervals of epilepsy patients. First of all, the multivariate autoregressive model was established for a moving window and the direction and intensity of information flow based on PDC analysis was calculated. Then, the outflow information related to certain EEG channel could be obtained by summing up the intensity of information flow propagated to other EEG channels in order to reduce the feature dimensionality. At last, according to the pathological features of epileptic seizures, the outflow information was regarded as the input vectors to a support vector machine classifier for discriminating interictal periods and ictal periods of EEG signals. The proposed method had achieved a good performance with the correct rate of 98.3%, the selectivity rate of 67.88%, the sensitivity rate of 91.44%, the specificity rate of 99.34%, and the average detection rate of 95.39%, which demonstrated that this method was suitable for detecting the seizure intervals of epilepsy patients. By comparing with other existing techniques, the proposed method based on PDC analysis achieved significant improvement in terms of seizure detection.

Automatic seizure detection using Stockwell transform and boosting algorithm for long-term EEG

Automatic detection of seizures has vital significance for epileptic diagnosis and can efficiently reduce the workload of the medical staff. In this study, a novel seizure detection method based on Stockwell transform is proposed for intracranial long-term EEG data. The Stockwell transform is employed to obtain the time-frequency representation of the EEG signals, and then the power spectral density is calculated in the time-frequency plane to characterize the behavior of EEG recordings. After that, a classifier based on gradient boosting algorithm is used to make the classification. Finally, the postprocessing is utilized on the outputs of the classifier to obtain more stable and accurate detection results, which includes Kalman filter, threshold judgment, and collar technique. The performance of this method is assessed on the publicly available EEG database which contains approximately 533h of intracranial EEG recordings. The experimental results indicate that the proposed method can achieve a satisfactory sensitivity of 94.26%, a specificity of 96.34%, as well as a very short delay time of 0.56s.

Detection of epileptiform activity in EEG signals based on time-frequency and non-linear analysis

We present a new technique for detection of epileptiform activity in EEG signals. After preprocessing of EEG signals we extract representative features in time, frequency and time-frequency domain as well as using non-linear analysis. The features are extracted in a few frequency sub-bands of clinical interest since these sub-bands showed much better discriminatory characteristics compared with the whole frequency band. Then we optimally reduce the dimension of feature space to two using scatter matrices. A decision about the presence of epileptiform activity in EEG signals is made by quadratic classifiers designed in the reduced two-dimensional feature space. The accuracy of the technique was tested on three sets of electroencephalographic (EEG) signals recorded at the University Hospital Bonn: surface EEG signals from healthy volunteers, intracranial EEG signals from the epilepsy patients during the seizure free interval from within the seizure focus and intracranial EEG signals of epileptic seizures also from within the seizure focus. An overall detection accuracy of 98.7% was achieved.

Genetic Programming and Frequent Itemset Mining to Identify Feature Selection Patterns of iEEG and fMRI Epilepsy Data

Pattern classification for intracranial electroencephalogram (iEEG) and functional magnetic resonance imaging (fMRI) signals has furthered epilepsy research toward understanding the origin of epileptic seizures and localizing dysfunctional brain tissue for treatment. Prior research has demonstrated that implicitly selecting features with a genetic programming (GP) algorithm more effectively determined the proper features to discern biomarker and non-biomarker interictal iEEG and fMRI activity than conventional feature selection approaches. However for each the iEEG and fMRI modalities, it is still uncertain whether the stochastic properties of indirect feature selection with a GP yield (a) consistent results within a patient data set and (b) features that are specific or universal across multiple patient data sets. We examined the reproducibility of implicitly selecting features to classify interictal activity using a GP algorithm by performing several selection trials and subsequent frequent itemset mining (FIM) for separate iEEG and fMRI epilepsy patient data. We observed within-subject consistency and across-subject variability with some small similarity for selected features, indicating a clear need for patient-specific features and possible need for patient-specific feature selection or/and classification. For the fMRI, using nearest-neighbor classification and 30 GP generations, we obtained over 60% median sensitivity and over 60% median selectivity. For the iEEG, using nearest-neighbor classification and 30 GP generations, we obtained over 65% median sensitivity and over 65% median selectivity except one patient.

Ensemble classifier for epileptic seizure detection for imperfect EEG data

Brain status information is captured by physiological electroencephalogram (EEG) signals, which are extensively used to study different brain activities. This study investigates the use of a new ensemble classifier to detect an epileptic seizure from compressed and noisy EEG signals. This noise-aware signal combination (NSC) ensemble classifier combines four classification models based on their individual performance. The main objective of the proposed classifier is to enhance the classification accuracy in the presence of noisy and incomplete information while preserving a reasonable amount of complexity. The experimental results show the effectiveness of the NSC technique, which yields higher accuracies of 90% for noiseless data compared with 85%, 85.9%, and 89.5% in other experiments. The accuracy for the proposed method is 80% when SNR=1 dB, 84% when SNR=5 dB, and 88% when SNR=10 dB, while the compression ratio (CR) is 85.35% for all of the datasets mentioned.

Transductive domain adaptive learning for epileptic electroencephalogram recognition

OBJECTIVE

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.

METHODS

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.

RESULTS

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.

CONCLUSION

The proposed transfer-learning-based method has better classification accuracy and adaptability than the conventional methods in classifying EEG signals for epilepsy detection.

The effect of multiscale PCA de-noising in epileptic seizure detection

In this paper we describe the effect of Multiscale Principal Component Analysis (MSPCA) de-noising method in terms of epileptic seizure detection. In addition, we developed a patient-independent seizure detection algorithm using Freiburg EEG database. Each patient contains datasets called "ictal" and "interictal". Window length of 16 s was applied to extract EEG segments from datasets of each patient. Furthermore, Power Spectral Density (PSD) of each EEG segment was estimated using different spectral analysis methods. Afterwards, these values were fed as input to different machine learning methods that were responsible for seizure detection. We also applied MSPCA de-noising method to EEG segments prior to PSD estimation to determine if MSPCA can further enhance the classifiers' performance. The MSPCA drastically improved both the sensitivity and the specificity, increasing the overall accuracy of all three classifiers up to 20%. The best overall detection accuracy (99.59%) was achieved when Eigenvector analysis was used for frequency estimation, and C4.5 as a classifier. The experiment results show that MSPCA is an effective de-noising method for improving the classification performance in epileptic seizure detection.

Identification and monitoring of brain activity based on stochastic relevance analysis of short-time EEG rhythms

Background

The extraction of physiological rhythms from electroencephalography (EEG) data and their automated analyses are extensively studied in clinical monitoring, to find traces of interictal/ictal states of epilepsy.

Methods

Because brain wave rhythms in normal and interictal/ictal events, differently influence neuronal activity, our proposed methodology measures the contribution of each rhythm. These contributions are measured in terms of their stochastic variability and are extracted from a Short Time Fourier Transform to highlight the non–stationary behavior of the EEG data. Then, we performed a variability–based relevance analysis by handling the multivariate short–time rhythm representation within a subspace framework. This maximizes the usability of the input information and preserves only the data that contribute to the brain activity classification. For neural activity monitoring, we also developed a new relevance rhythm diagram that qualitatively evaluates the rhythm variability throughout long time periods in order to distinguish events with different neuronal activities.

Results

Evaluations were carried out over two EEG datasets, one of which was recorded in a noise–filled environment. The method was evaluated for three different classification problems, each of which addressed a different interpretation of a medical problem. We perform a blinded study of 40 patients using the support–vector machine classifier cross–validation scheme. The obtained results show that the developed relevance analysis was capable of accurately differentiating normal, ictal and interictal activities.

Conclusions

The proposed approach provides the reliable identification of traces of interictal/ictal states of epilepsy. The introduced relevance rhythm diagrams of physiological rhythms provides effective means of monitoring epileptic seizures; additionally, these diagrams are easily implemented and provide simple clinical interpretation. The developed variability–based relevance analysis can be translated to other monitoring applications involving time–variant biomedical data.

Patient specific seizure prediction system using Hilbert spectrum and Bayesian networks classifiers

The aim of this paper is to develop an automated system for epileptic seizure prediction from intracranial EEG signals based on Hilbert-Huang transform (HHT) and Bayesian classifiers. Proposed system includes decomposition of the signals into intrinsic mode functions for obtaining features and use of Bayesian networks with correlation based feature selection for binary classification of preictal and interictal recordings. The system was trained and tested on Freiburg EEG database. 58 hours of preictal data, 40-minute data blocks prior to each of 87 seizures collected from 21 patients, and 503.1 hours of interictal data were examined resulting in 96.55% sensitivity with 0.21 false alarms per hour, 13.896% average proportion of time spent in warning, and 33.21 minutes of average detection latency using 30-second EEG segments with 50% overlap and a simple postprocessing technique resulting in a decision (a seizure is expected/not expected) every 5 minutes. High sensitivity and low false positive rate with reasonable detection latency show that HHT based features are acceptable for patient specific seizure prediction from intracranial EEG data. Time spent for testing an EEG segment was 4.1451 seconds on average, which makes the system viable for use in real-time seizure control systems.

Performance-power consumption tradeoff in wearable epilepsy monitoring systems

Automated seizure detection methods can be used to reduce time and costs associated with analyzing large volumes of ambulatory EEG recordings. These methods however have to rely on very complex, power hungry algorithms, implemented on the system backend, in order to achieve acceptable levels of accuracy. In size, and therefore power-constrained EEG systems, an alternative approach to the problem of data reduction is online data selection, in which simpler algorithms select potential epileptiform activity for discontinuous recording but accurate analysis is still left to a medical practitioner. Such a diagnostic decision support system would still provide doctors with information relevant for diagnosis while reducing the time taken to analyze the EEG. For wearable systems with limited power budgets, data selection algorithm must be of sufficiently low complexity in order to reduce the amount of data transmitted and the overall power consumption. In this paper, we present a low-power hardware implementation of an online epileptic seizure data selection algorithm with encryption and data transmission and demonstrate the tradeoffs between its accuracy and the overall system power consumption. We demonstrate that overall power savings by data selection can be achieved by transmitting less than 40% of the data. We also show a 29% power reduction when selecting and transmitting 94% of all seizure events and only 10% of background EEG.

Detection of seizure and epilepsy using higher order statistics in the EMD domain

In this paper, a method using higher order statistical moments of EEG signals calculated in the empirical mode decomposition (EMD) domain is proposed for detecting seizure and epilepsy. The appropriateness of these moments in distinguishing the EEG signals is investigated through an extensive analysis in the EMD domain. An artificial neural network is employed as the classifier of the EEG signals wherein these moments are used as features. The performance of the proposed method is studied using a publicly available benchmark database for various classification cases that include healthy, interictal (seizure-free interval) and ictal (seizure), healthy and seizure, nonseizure and seizure, and interictal and ictal, and compared with that of several recent methods based on time-frequency analysis and statistical moments. It is shown that the proposed method can provide, in almost all the cases, 100% accuracy, sensitivity, and specificity, especially in the case of discriminating seizure activities from the nonseizure ones for patients with epilepsy while being much faster as compared to the time-frequency analysis-based techniques.

Epileptic seizure classification in EEG signals using second-order difference plot of intrinsic mode functions

Epilepsy is a neurological disorder which is characterized by transient and unexpected electrical disturbance of the brain. The electroencephalogram (EEG) is a commonly used signal for detection of epileptic seizures. This paper presents a new method for classification of ictal and seizure-free EEG signals. The proposed method is based on the empirical mode decomposition (EMD) and the second-order difference plot (SODP). The EMD method decomposes an EEG signal into a set of symmetric and band-limited signals termed as intrinsic mode functions (IMFs). The SODP of IMFs provides elliptical structure. The 95% confidence ellipse area measured from the SODP of IMFs has been used as a feature in order to discriminate seizure-free EEG signals from the epileptic seizure EEG signals. The feature space obtained from the ellipse area parameters of two IMFs has been used for classification of ictal and seizure-free EEG signals using the artificial neural network (ANN) classifier. It has been shown that the feature space formed using ellipse area parameters of first and second IMFs has given good classification performance. Experimental results on EEG database available by the University of Bonn, Germany, are included to illustrate the effectiveness of the proposed method.

Feature extraction with stacked autoencoders for epileptic seizure detection

Scalp electroencephalogram (EEG), a recording of the brain's electrical activity, has been used to diagnose and detect epileptic seizures for a long time. However, most researchers have implemented seizure detectors by manually hand-engineering features from observed EEG data, and used them in seizure detection, which might not scale well to new patterns of seizures. In this paper, we investigate the possibility of utilising unsupervised feature learning, the recent development of deep learning, to automatically learn features from raw, unlabelled EEG data that are representative enough to be used in seizure detection. We develop patient-specific seizure detectors by using stacked autoencoders and logistic classifiers. A two-step training consisting of the greedy layer-wise and the global fine-tuning was used to train our detectors. The evaluation was performed by using labelled dataset from the CHB-MIT database, and the results showed that all of the test seizures were detected with a mean latency of 3.36 seconds, and a low false detection rate.

Teager Energy Based Filter-Bank Cepstra in EEG Classification for Seizure Detection Using Radial Basis Function Neural Network

About 1–3% of the world population suffers from epilepsy. Epileptic seizures are abnormal sudden discharges in the brain with signatures manifesting in the electroencephalograph (EEG) recordings by frequency changes and increased amplitudes. These changes, in this work, are captured through static and dynamic features derived from three Teager energy based filter-bank cepstra (TE-FB-CEPs). We compared the performance of linear, logarithmic, and Mel frequency scale TE-FB-CEPs using radial basis function neural network in general epileptic seizure detection. The comparison is tried on eight different classification problems which encompass all the possible discriminations in the medical field related to epilepsy. In a previous study, using traditional cepstrum on the same database, we had found that the composite vectors showed a degraded performance in seizure detection. In this study, however, irrespective of frequency scaling used, it is found that the composite vectors of TE-FB-CEPs maintain excellent overall accuracy in all the eight classification problems.

A new parametric feature descriptor for the classification of epileptic and control EEG records in pediatric population

This study evaluates the sensitivity, specificity and accuracy in associating scalp EEG to either control or epileptic patients by means of artificial neural networks (ANNs) and support vector machines (SVMs). A confluence of frequency and temporal parameters are extracted from the EEG to serve as input features to well-configured ANN and SVM networks. Through these classification results, we thus can infer the occurrence of high-risk (epileptic) as well as low risk (control) patients for potential follow up procedures.

Adaptive Parametric Spectral Estimation with Kalman Smoothing for Online Early Seizure Detection

Tracking spectral changes in neural signals, such as local field potentials (LFPs) and scalp or intracranial electroencephalograms (EEG, iEEG), is an important problem in early detection and prediction of seizures. Most approaches have focused on either parametric or nonparametric spectral estimation methods based on moving time windows. Here, we explore an adaptive (time-varying) parametric ARMA approach for tracking spectral changes in neural signals based on the fixed-interval Kalman smoother. We apply the method to seizure detection based on spectral features of intracortical LFPs recorded from a person with pharmacologically intractable focal epilepsy. We also devise and test an approach for real-time tracking of spectra based on the adaptive parametric method with the fixed-interval Kalman smoother. The order of ARMA models is determined via the AIC computed in moving time windows. We quantitatively demonstrate the advantages of using the adaptive parametric estimation method in seizure detection over nonparametric alternatives based exclusively on moving time windows. Overall, the adaptive parametric approach significantly improves the statistical separability of interictal and ictal epochs.

A boosted cascade for efficient epileptic seizure detection

Seizure detection from electroencephalogram (EEG) plays an important role for epilepsy therapy. Due to the diversity of seizure EEG patterns between different individuals, multiple features are necessary for high accuracy since a single feature could hardly encode all types of epileptiform discharges. However, a large feature set inevitably causes the increase of the computational cost. This paper proposes a boosted cascade chain to obtain both high detection performance and high computational efficiency. Sixteen features that are widely used in seizure detection are implemented. Considering the sequential characteristics of EEG signals, the features are extracted on each 1-second segment and its former three segments. Thus, a total of 64 features are used to construct a feature pool. Based on the feature pool, Real AdaBoost is used to select a group of effective features, on which weak classifiers are learned to assemble a strong classifier. The strong classifier is transformed to a cascade classifier by reordering the weak classifiers and learning a threshold for each weak classifier. The cascade classifier still has the similar classification strength to the original strong classifier. More importantly, it is able to reject easy non-seizure samples by the first a few weak classifiers in the cascade, thus high computational efficiency can be obtained. To evaluate our method, 90.6-hour EEG signals from four patients are tested. The experimental results show that our method can achieve an average accuracy of 95.31% and an average detection rate of 91.29% with the false positive rate of 4.68%. On average, only about 4 features are used. Compared with support vector machine (SVM), our method is much more efficient with the similar detection performance.

High-Performance Seizure Detection System Using a Wavelet-Approximate Entropy-fSVM Cascade With Clinical Validation

The classification of electroencephalography (EEG) signals is one of the most important methods for seizure detection. However, verification of an atypical epileptic seizure often can only be done through long-term EEG monitoring for 24 hours or longer. Hence, automatic EEG signal analysis for clinical screening is necessary for the diagnosis of epilepsy. We propose an EEG analysis system of seizure detection, based on a cascade of wavelet-approximate entropy for feature selection, Fisher scores for adaptive feature selection, and support vector machine for feature classification. Performance of the system was tested on open source data, and the overall accuracy reached 99.97%. We further tested the performance of the system on clinical EEG obtained from a clinical EEG laboratory and bedside EEG recordings. The results showed an overall accuracy of 98.73% for routine EEG, and 94.32% for bedside EEG, which verified the high performance and usefulness of such a cascade system for seizure detection. Also, the prediction model, trained by routine EEG, can be successfully generalized to bedside EEG of independent patients.

Detection of artifacts from high energy bursts in neonatal EEG

Detection of non-cerebral activities or artifacts, intermixed within the background EEG, is essential to discard them from subsequent pattern analysis. The problem is much harder in neonatal EEG, where the background EEG contains spikes, waves, and rapid fluctuations in amplitude and frequency. Existing artifact detection methods are mostly limited to detect only a subset of artifacts such as ocular, muscle or power line artifacts. Few methods integrate different modules, each for detection of one specific category of artifact. Furthermore, most of the reference approaches are implemented and tested on adult EEG recordings. Direct application of those methods on neonatal EEG causes performance deterioration, due to greater pattern variation and inherent complexity. A method for detection of a wide range of artifact categories in neonatal EEG is thus required. At the same time, the method should be specific enough to preserve the background EEG information. The current study describes a feature based classification approach to detect both repetitive (generated from ECG, EMG, pulse, respiration, etc.) and transient (generated from eye blinking, eye movement, patient movement, etc.) artifacts. It focuses on artifact detection within high energy burst patterns, instead of detecting artifacts within the complete background EEG with wide pattern variation. The objective is to find true burst patterns, which can later be used to identify the Burst-Suppression (BS) pattern, which is commonly observed during newborn seizure. Such selective artifact detection is proven to be more sensitive to artifacts and specific to bursts, compared to the existing artifact detection approaches applied on the complete background EEG. Several time domain, frequency domain, statistical features, and features generated by wavelet decomposition are analyzed to model the proposed bi-classification between burst and artifact segments. A feature selection method is also applied to select the feature subset producing highest classification accuracy. The suggested feature based classification method is executed using our recorded neonatal EEG dataset, consisting of burst and artifact segments. We obtain 78% sensitivity and 72% specificity as the accuracy measures. The accuracy obtained using the proposed method is found to be about 20% higher than that of the reference approaches. Joint use of the proposed method with our previous work on burst detection outperforms reference methods on simultaneous burst and artifact detection. As the proposed method supports detection of a wide range of artifact patterns, it can be improved to incorporate the detection of artifacts within other seizure patterns and background EEG information as well.

A physiology-based seizure detection system for multichannel EEG

Background

Epilepsy is a common chronic neurological disorder characterized by recurrent unprovoked seizures. Electroencephalogram (EEG) signals play a critical role in the diagnosis of epilepsy. Multichannel EEGs contain more information than do single-channel EEGs. Automatic detection algorithms for spikes or seizures have traditionally been implemented on single-channel EEG, and algorithms for multichannel EEG are unavailable.

Methodology

This study proposes a physiology-based detection system for epileptic seizures that uses multichannel EEG signals. The proposed technique was tested on two EEG data sets acquired from 18 patients. Both unipolar and bipolar EEG signals were analyzed. We employed sample entropy (SampEn), statistical values, and concepts used in clinical neurophysiology (e.g., phase reversals and potential fields of a bipolar EEG) to extract the features. We further tested the performance of a genetic algorithm cascaded with a support vector machine and post-classification spike matching.

Principal Findings

We obtained 86.69% spike detection and 99.77% seizure detection for Data Set I. The detection system was further validated using the model trained by Data Set I on Data Set II. The system again showed high performance, with 91.18% detection of spikes and 99.22% seizure detection.

Conclusion

We report a de novo EEG classification system for seizure and spike detection on multichannel EEG that includes physiology-based knowledge to enhance the performance of this type of system.

Automated diagnosis of epilepsy using EEG power spectrum

Interictal electroencephalography (EEG) has clinically meaningful limitations in its sensitivity and specificity in the diagnosis of epilepsy because of its dependence on the occurrence of epileptiform discharges. We have developed a computer-aided diagnostic (CAD) tool that operates on the absolute spectral energy of the routine EEG and has both substantially higher sensitivity and negative predictive value than the identification of interictal epileptiform discharges. Our approach used a multilayer perceptron to classify 156 patients admitted for video-EEG monitoring. The patient population was diagnostically diverse; 87 were diagnosed with either generalized or focal seizures. The remainder of the patients were diagnosed with nonepileptic seizures. The sensitivity was 92% (95% confidence interval [CI] 85-97%) and the negative predictive value was 82% (95% CI 67-92%). We discuss how these findings suggest that this CAD can be used to supplement event-based analysis by trained epileptologists.

Automatic seizure detection in SEEG using high frequency activities in wavelet domain

Existing automatic detection techniques show high sensitivity and moderate specificity, and detect seizures a relatively long time after onset. High frequency (80-500 Hz) activity has recently been shown to be prominent in the intracranial EEG of epileptic patients but has not been used in seizure detection. The purpose of this study is to investigate if these frequencies can contribute to seizure detection. The system was designed using 30 h of intracranial EEG, including 15 seizures in 15 patients. Wavelet decomposition, feature extraction, adaptive thresholding and artifact removal were employed in training data. An EMG removal algorithm was developed based on two features: Lack of correlation between frequency bands and energy-spread in frequency. Results based on the analysis of testing data (36 h of intracranial EEG, including 18 seizures) show a sensitivity of 72%, a false detection of 0.7/h and a median delay of 5.7 s. Missed seizures originated mainly from seizures with subtle or absent high frequencies or from EMG removal procedures. False detections were mainly due to weak EMG or interictal high frequency activities. The system performed sufficiently well to be considered for clinical use, despite the exclusive use of frequencies not usually considered in clinical interpretation. High frequencies have the potential to contribute significantly to the detection of epileptic seizures.