Normal EEG Signals Research Articles

A novel epilepsy detection approach using intrinsic multiscale entropy analysis and DSEAM-enhanced 1D-ResNets

Abstract Epilepsy, a prevalent neurological disorder, typically requires a complex diagnostic process involving medical history inquiry, physical examination, head computed tomography, and electroencephalogram (EEG) visual examination. Among the existing epilepsy automated detection algorithms, machine learning methods require manual feature selection. Most of deep learning algorithms for automatic detection have high complexity and computational complexity. To address this issue, this study proposes a high-precision, robust, and low computational epilepsy automatic detection algorithm based on EEG signal processing. The study utilizes ensemble empirical mode decomposition to preprocess the original EEG signal, breaking it down into intrinsic mode functions (IMFs) across various frequency bands. These IMFs contain information about epilepsy occurrence within the signal at different frequency bands. To enhance computational efficiency and reduce data dimension, the refined composite multiscale dispersion entropy of each IMF is further computed at different scales, referred to as intrinsic multiscale entropy (IME) analysis. IME analysis consolidates epilepsy occurrence information from EEG signals across different frequency bands and scales, linking entropy values to generate feature vectors. Drawing inspiration from successful deep residual networks and Squeeze-and-Excitation networks, the study introduces a double Squeeze-and-Excitation attention module enhanced one-dimensional residual network to classify one-dimensional feature vectors. The proposed method was tested on epilepsy dataset from University of Bonn, and the results demonstrated superior classification performance. In the experiment, the distinction between normal and epileptic EEG signals achieved 100% accuracy rate, while distinguishing between normal, epileptic interval, and epileptic EEG signals achieved accuracy rate of 99.41%.

Automatic detection of epileptic seizure based on one dimensional cascaded convolutional autoencoder with adaptive window-thresholding

Objective. Identifying the seizure occurrence period (SOP) in extended EEG recordings is crucial for neurologists to diagnose seizures effectively. However, many existing computer-aided diagnosis systems for epileptic seizure detection (ESD) primarily focus on distinguishing between ictal and interictal states in EEG recordings. This focus has limited their application in clinical settings, as these systems typically rely on supervised learning approaches that require labeled data.Approach. To address this, our study introduces an unsupervised learning framework for ESD using a 1D- cascaded convolutional autoencoder (1D-CasCAE). In this approach, EEG recordings from selected patients in the CHB-MIT datasets are first segmented into 5 s epochs. Eight informative channels are chosen based on the correlation coefficient and Shannon entropy. The 1D-CasCAE is designed to autonomously learn the characteristic patterns of interictal (non-seizure) segments through downsampling and upsampling processes. The integration of adaptive thresholding and a moving window significantly enhances the model's robustness, enabling it to accurately identify ictal segments in long EEG recordings.Main results. Experimental results demonstrate that the proposed 1D-CasCAE effectively learns normal EEG signal patterns and efficiently detects anomalies (ictal segments) using reconstruction errors. When compared with other leading methods in anomaly detection, our model exhibits superior performance, as evidenced by its average Gmean, sensitivity, specificity, precision, and false positive rate scores of 98.00% ± 3.51%, 94.94% ± 6.92%, 99.60% ± 0.30%, 79.92% ± 13.56% and 0.0044 ± 0.0030 h-1respectively for a typical patient in CHB-MIT datasets.Significance. The developed model framework can be employed in clinical settings, replacing the manual inspection process of EEG signals by neurologists. Furthermore, the proposed automated system can adapt to each patient's SOP through the use of variable time windows for seizure detection.

An autoencoder model based on one-dimensional neural network for epileptic EEG anomaly detection

We propose an autoencoder model based on a one-dimensional convolutional neural network (1DCNN) as the feature extraction network for efficient detection of epileptic EEG anomalies. The local information of normal EEG signals was captured by utilizing the local feature extraction ability of 1DCNN for training of an autoencoder to learn the expression of normal EEG data in low dimensional feature space. With the difference between the input and output as the anomaly score, the threshold was determined by the optimal equilibrium point of the ROC curve, and the EEG signals exceeding the threshold were diagnosed as the seizure data. The performance of the 1DCNN-AE epilepsy detection model was evaluated using the publicly available CHB-MIT scalp EEG dataset and TUH scalp EEG dataset. The AUC of the 1DCNN-AE model reached 0.890 of CHB-MIT and 0.686 of TUH under the average level of patients, and the epilepsy detection rate reached 0.974 and 0.893, and these results were better than the latest epilepsy anomaly detection models LSTM-VAE and GRU-VAE. The 1DCNN model had a parameter quantity of 58.5M, which was at the same level with LSTM-VAE (47.4 M) and GRU-VAE (36.9 M) but with much smaller FLOPs (0.377 G) than LSTM-VAE (21.6 G) and GRU-VAE (16.2 G). The autoencoder model based one-dimensional convolutional neural network can effectively detect abnormal EEG signals in epileptic seizure.

AI-enhanced EEG signal interpretation: A novel approach using texture analysis with random forests

We hypothesize that the Gray-Level Co-occurrence Matrix (GLCM) and the Run-Length Matrix (RLM) techniques can effectively quantify discrete changes in EEG signals, and that the features extracted from these matrices can be utilized to train a Random Forest (RF) model. Our contribution includes the development of a robust code in sci-kit learn for a hypothetical model that, after adequate training and testing, could be used to detect and remove artifacts as well as differentiate between physiological and pathological EEG signals. Moreover, our approach envisions the RF model as a powerful tool capable of differentiating between normal and abnormal EEG signals. This approach could lead to the development of more potent AI tools that enhance clinical decision-making in neurology and psychiatry.

Seizure detection using nonlinear measures over EEG frequency bands and deep learning classifiers

Epilepsy is a brain disorder that causes patients to suffer from convulsions, which affects their behavior and way of life. Epilepsy can be detected with electroencephalograms (EEGs), which record brain neural activity. Traditional approaches for detecting epileptic seizures from an EEG signal are time-consuming and annoying. To supersede these traditional methods, a myriad of automated seizure detection frameworks based on machine learning techniques have recently been developed. Feature extraction and classification are the two essential phases for seizure detection. The classifier assigns the proper class label after feature extraction lowers the input pattern space while maintaining useful features. This paper proposes a new feature extraction method based on calculating nonlinear features from the most relevant EEG frequency bands. The EEG signal is first decomposed into smaller time segments from which a vector of nonlinear features is computed and supplied to machine learning (ML) and deep learning (DL) classifiers. Experiments on the Bonn dataset reveals an accuracy of 99.7% reached in classifying normal and ictal EEG signals; and an accuracy of 98.8% in the discrimination of ictal and interictal EEG signals. Furthermore, a performance of 100% is achieved on the Hauz Khas dataset. The classification results of the proposed approach were compared to those from published state of the art techniques. Our results are equivalent to or better than some recent studies appeared in the literature.

Advanced framework for epilepsy detection through image-based EEG signal analysis.

Recurrent and unpredictable seizures characterize epilepsy, a neurological disorder affecting millions worldwide. Epilepsy diagnosis is crucial for timely treatment and better outcomes. Electroencephalography (EEG) time-series data analysis is essential for epilepsy diagnosis and surveillance. Complex signal processing methods used in traditional EEG analysis are computationally demanding and difficult to generalize across patients. Researchers are using machine learning to improve epilepsy detection, particularly visual feature extraction from EEG time-series data. This study examines the application of a Gramian Angular Summation Field (GASF) approach for the analysis of EEG signals. Additionally, it explores the utilization of image features, specifically the Scale-Invariant Feature Transform (SIFT) and Oriented FAST and Rotated BRIEF (ORB) techniques, for the purpose of epilepsy detection in EEG data. The proposed methodology encompasses the transformation of EEG signals into images based on GASF, followed by the extraction of features utilizing SIFT and ORB techniques, and ultimately, the selection of relevant features. A state-of-the-art machine learning classifier is employed to classify GASF images into two categories: normal EEG patterns and focal EEG patterns. Bern-Barcelona EEG recordings were used to test the proposed method. This method classifies EEG signals with 96% accuracy using SIFT features and 94% using ORB features. The Random Forest (RF) classifier surpasses state-of-the-art approaches in precision, recall, F1-score, specificity, and Area Under Curve (AUC). The Receiver Operating Characteristic (ROC) curve shows that Random Forest outperforms Support Vector Machine (SVM) and k-Nearest Neighbors (k-NN) classifiers. The suggested method has many advantages over time-series EEG data analysis and machine learning classifiers used in epilepsy detection studies. A novel image-based preprocessing pipeline using GASF for robust image synthesis and SIFT and ORB for feature extraction is presented here. The study found that the suggested method can accurately discriminate between normal and focal EEG signals, improving patient outcomes through early and accurate epilepsy diagnosis.

FBSE-based Approach for Discriminating Seizure and Normal EEG Signals

FBSE-based Approach for Discriminating Seizure and Normal EEG Signals

EEG-Based Depression Detection in the Prefrontal Cortex Lobe using mRMR Feature Selection and Bidirectional LSTM

Depression can induce significant anguish and impair one's ability to perform effectively in professional, academic, and familial settings. This condition has the potential to result in suicide. Annually, the number of deaths resulting from suicide exceeds 700,000. Among individuals aged 15-29, suicide has emerged as the fourth most prevalent cause of mortality. Challenges in treating depression include limited accessibility to mental health care in rural regions and misdiagnosis resulting from subjective evaluations, wherein insufficient expertise can contribute to inaccurate diagnoses. Electroencephalography (EEG) has gained popularity as a tool for the identification and study of a number of mental illnesses in the past several years. Therefore, an automated technique is required to precisely distinguish between normal EEG signals and depression signals. This research focused on developing an EEG-based depression detection system in the prefrontal cortex lobe area (Fp1, Fpz, and Fp2). One of the developments carried out in this research is the implementation of Bidirectional Long Short-Term Memory (Bi-LSTM) as the model classification and minimum redundancy maximum relevance (mRMR) feature selection. Results suggest that the combination of mRMR feature selection with 25 features and the Bidirectional LSTM obtained 92.83% for the accuracy.

Classification of EEG Signal using Independent Component Analysis and Discrete Wavelet Transform based on Linear Discriminant Analysis

Autism Spectrum Disorder (ASD) is a neurodevelopment syndrome decreasing sufferers' social interaction, communication skills, and emotional expression. Autism syndrome can be detected using an electroencephalogram (EEG). This study utilized the EEG of autistic people to support the classification study of machine learning schemes to produce the best accuracy. One of the best approaches to classify the EEG signal is The Linear Discriminant Analysis (LDA), a machine learning technique to classify autism and normal EEG signals. LDA was chosen because it can maximize the distance between classes and minimize the number of scatters by utilizing between and within-class functions. This method was combined with other methods: Independent Components Analysis (ICA) and Discrete Wavelet Transform (DWT), to improve the accuracy system. ICA removes artifacts or signals other than brain signals that can cause noise in the EEG signal, so the analyzed signal was a complete EEG signal without other factors. DWT can help increase noise suppression in the EEG signal and provide signal information through frequency and time representation. The EEG dataset was collated from 16 children (eight autistic and eight normal). The signals from the dataset were filtered by artifacts using ICA, decomposed by three levels through DWT, and classified using the Linear Discriminant Analysis (LDA) technique. Using the Confusion Matrix, the results reveal the best accuracy of 99%.

The effect of power spectral density on the electroencephalography of autistic children based on the welch periodogram method

Autism spectrum disorder (ASD) is a serious mental disorder affecting social behavior. Some children also face intellectual delay. In people with ASD, the signals detected have abnormalities compared to normal people. This can be a reference in diagnosing the disorder with electroencephalography (EEG). This study will analyze the effect of Power spectral density (PSD) on the EEG of autistic children and also compare it with the PSD value on the EEG of normal children using the Welch Periodogram method approach. In the preprocessing stage, the Independent Component Analysis (ICA) method will be applied to remove artifacts, and a Finite Impulse Response (FIR) filter to reduce noise in the EEG signal. The study results indicate differences in the PSD values ​​obtained in the autistic and normal EEG signals. The PSD value obtained in the autistic EEG signal is higher than the normal EEG signal in all frequency sub-bands. From the study results, the highest PSD value obtained by the autistic EEG signal is in the delta sub-band, which is 54.06 dB/Hz, while the normal EEG signal is only 33.14 dB/Hz at the same frequency sub-band. And in the Alpha and Beta sub-bands, the normal EEG signal increases the PSD value, while in the autistic EEG signal, the PSD value decreases in the Alpha and Beta sub-bands. In addition, FIR and ICA methods can also reduce noise and artifacts contained in autistic and normal EEG signals.

An Efficient Method for Classification of Alcoholic and Normal Electroencephalogram Signals Based on Selection of an Appropriate Feature.

Alcohol addiction contributes to disorders in brain's normal patterns. Analysis of electroencephalogram (EEG) signal helps to diagnose and classify alcoholic and normal EEG signal. One-second EEG signal was applied to classify alcoholic and normal EEG signal. To determine discriminative feature and EEG channel between the alcoholic and normal EEG signal, different frequency and non-frequency features of EEG signal, including power of EEG signal, permutation entropy (PE), approximate entropy (ApEn), katz fractal dimension (katz FD) and Petrosion fractal dimension (Petrosion FD) were extracted from alcoholic and normal EEG signal. Statistical analysis and Davis-Bouldin criterion (DB) were utilized to specify and select most discriminative feature and EEG channel between the alcoholic and normal EEG signal. Results of statistical analysis and DB criterion showed that the Katz FD in FP2 channel showed the best discrimination between the alcoholic and normal EEG signal. The Katz FD in FP2 channel showed the accuracies of 98.77% and 98.5% by two classifiers with 10-fold cross validation. This method helps to diagnose alcoholic and normal EEG signal with the minimum number of feature and channel, which provides low computational complexity. This is helpful to faster and more accurate classification of normal and alcoholic subjects.

A novel model based on a 1D-ResCNN and transfer learning for processing EEG attenuation

EEG signals are valuable signals in clinical medicine, brain research, and the study of neurological illnesses. However, EEG signal attenuation may occur at any time from signal generation through BCI device acquisition due to defects in the brain–computer interface (BCI) devices, restrictions in the dynamic network, and individual variations across the subjects. The attenuation of EEG data will alter the data distribution and lead to information fuzziness, substantially influencing subsequent EEG research. A model based on one-dimensional residual convolutional neural networks (1D-ResCNN) and transfer learning is proposed in this article to reduce the negative impacts of EEG attenuation. An end-to-end manner maps an attenuated EEG signal to a normal EEG signal. The structure employs a multi-level residual connection structure with varying weight coefficients, transferring characteristics from the bottom to the top of the convolutional neural network, enhancing feature learning. In addition, we initialize the subsequent denoising model using the transfer learning method. The combination of these two networks can well solve the attenuation problem of EEG signals. Experiments are carried out using the EEG-denoisenet data set. According to the findings, the model can yield a clear waveform with a decent SNR and RRMSE value.

Psychological distance and user engagement in online exhibitions: Visualization of moiré patterns based on electroencephalography signals.

The COVID-19 pandemic has significantly affected the exhibition of artworks in museums and galleries. Many have displayed their collection online. In this context, experiencing an online exhibition is essential for visitors to appreciate and understand the artwork. Compared with offline exhibitions, visitors to online exhibitions are often unable to communicate their experiences with other visitors. Therefore, in this study, by facilitating communication via Zoom call, we established a system that allows two people to visit the museum together through the Google Arts and Culture (GA&C) website. To reduce the psychological distance between online visitors and help increase user engagement, we designed and developed a media device based on moiré pattern visualization of electroencephalography (EEG) signals. The participants were divided into two groups to visit the online museum, communicating remotely through Zoom. The objective of this study was to test whether a real-time EEG signal visualization device could help close the psychological distance between participants and whether it could increase user engagement with the online exhibition. Participants were randomly assigned to either the normal online exhibition experience (NOEE) group or EEG signal visualization device (ESVD) group. Participants in the NOEE group experienced four online exhibitions (Task1, Task2, Task3, and Task4) together (two participants per test unit). The conditions for participants in the ESVD group remained the same, apart from adding a media device to enable them to visualize EEG signals. A total of 40 university students participated in this study. Independent samples t-tests revealed that participants in the ESVD group perceived a significantly closer psychological distance between themselves and the participants on the opposite side than those in the NOEE group (t = −2.699; p = 0.008 < 0.05). A one-way ANOVA revealed that participants experienced Task3 with significantly closer psychological distance assessments than Task1 (p = 0.002 < 0.05), Task2 (p = 0.000 < 0.05), and Task4 (p = 0.001 < 0.05). Repeated ANOVAs revealed that participants in the ESVD group had higher overall user engagement than those in the NOEE group, with marginal significance (p = 0.056 < 0.1). Thus, this study shows that EEG visualization media devices can reduce the psychological distance between two participants when experiencing an online exhibition. Moreover, it can increase user engagement to some extent.

Identification of normal and depression EEG signals in variational mode decomposition domain.

Early detection of depression is critical in assisting patients in receiving the best therapy possible to avoid negative repercussions. Depression detection using electroencephalogram (EEG) signals is a simple, low-cost, convenient, and accurate approach. This paper proposes a six-stage novel method for detecting depression using EEG signals. First, EEG signals are recorded from 44 subjects, with 22 subjects being normal and 22 subjects being depressed. Second, a simple notch filter with EEG signals differencing approach is employed for effective preprocessing. Third, the variational mode decomposition (VMD) approach is implemented for nonlinear and non-stationary EEG signals analysis, resulting in many modes. Fourth, mutual information-based novel modes selection criterion is proposed to select the most informative modes. In the fifth step, a combination of linear and nonlinear features are extracted from selected modes and at last, classification is performed with neural networks. In this study, a novel single feature is also proposed, which is made using Log energy, norm entropies and fluctuation index, which delivers 100% classification accuracy, sensitivity and specificity. By using these features, a novel depression diagnostic index is also proposed. This integrated index would assist in quicker and more objective identification of normal and depression EEG signals. The proposed computerized framework and the DDI can help health workers, large enterprises, and product developers build a real-time system.

Alcoholic EEG signals recognition based on phase space dynamic and geometrical features

Alcoholism is a severe disorder that leads to brain problems and associated cognitive, emotional and behavioral impairments. This disorder is typically diagnosed by a questionnaire technique known as CAGE that measures the severity of a drinking problem. This is a time-consuming, onerous, error-prone, and biased method. Hence, this article aims to establish a novel framework for automatic detecting alcoholism using electroencephalogram (EEG) signals, which can mitigate these issues and help clinicians and researchers. In the proposed framework, firstly, we explore the phase space dynamic of EEG signals for visualizing the chaotic nature and complexity of EEG signals. Secondly, we discover thirty-four graphical features for decoding the chaotic behavior of normal and alcoholic EEG signals. After that, we investigate thirteen feature selection in combination with eleven machine learning and neural network classifiers to select the best combination for the development of an efficient framework. The experimental results reveal that the proposed method achieves the highest classification performance involving 99.16% accuracy, 100% sensitivity and 98.36% specificity for the twenty-three features selected by Henry gas solubility optimization with feedforward neural network (FFNN). The proposed system provides a new visual biomarker for alcoholic detection. In addition, we developed two new indexes using clinically relevant features to distinguish normal and alcoholic classes with a single number. These indexes can help medical teams, commercial users as well as product developers to develop a real-time system.

EEG-based classification of normal and seizure types using relaxed local neighbour difference pattern and artificial neural network

Diagnosis of epileptic seizure types is vital for the neurologists to comprehend the cortical connectivity of the brain, and to initiate the apt treatment for the epileptic subjects at the earliest. However, the conventional method of manually analysing the EEG signals, for seizure diagnosis and its types identification is onerous and it is error-prone. Hence, a novel relaxed local neighbour difference pattern (RLNDiP) technique integrated with artificial neural network is introduced in this article for robust and accurate automatic classification of normal(N), generalized(G), and focal(F) EEG signals. The proposed unified framework consists of two phases. Firstly, the EEG signals are pre-processed and transformed into RLNDiP domain, using the proposed method. The histogram features are computed to perform the classification task. Secondly, the most prominent features are selected using the Kruskal–Wallis test and given into ANN network for discrimination. To find out the best number of hidden layers and hidden neurons of ANN, the experiments with 5-fold cross validation technique is conducted with the train set. Finally, with the best values ANN network is trained employing the entire train set and evaluated on test set. Proposed methodology is validated using the Karunya Institute of Technology and Sciences (KITS) database consisting of 86 EEG signals in each group. The proposed method attained an overall classification accuracy of 100% with train set, and 95.83% with test set. The results encourage the plausible widening of proposed methodology to real time ambience for assisting the neurologists in making efficient diagnosis of epileptic seizure types.

AUTOMATED CLASSIFICATION OF AUTISM SPECTRUM DISORDER USING EEG SIGNALS AND CONVOLUTIONAL NEURAL NETWORKS

Children suffering from Autism Spectrum Disorder (ASD) have impaired social communication, interaction and restricted and repetitive behaviors. ASD is caused by abnormal brain developments which give rise to the behavioral characteristics associated with ASD. The clinical diagnosis of ASD is performed on the basis of behavioral assessment and it causes a time delay in early intervention, as there is a time gap between abnormal brain developments and associated behavioral characteristics. Electroencephalography (EEG) is a technique which measures the electrical activity produced by the brain and it has been used to detect several neurological disorders. Studies have shown that there is a variation in the EEG signals of a normal subject and EEG signals of ASD subjects. In this study, we obtained scalograms of EEG signals by using Continuous Wavelet Transform (CWT). Pre-trained deep Convolutional Neural Networks (CNNs) such as GoogLeNet, AlexNet, MobileNet and SqueezeNet were used for extracting the features from scalograms and classification of obtained scalograms from EEG signals of normal and ASD subjects. We also used Support Vector Machine (SVM) algorithm and Relevance Vector Machine (RVM) for classification of the features extracted by the deep CNNs. The GoogLeNet, AlexNet, MobileNet and SqueezeNet deep CNNs achieved a validation accuracy of 75%, 75.84%, 79.45% and 82.98% in classifying the scalograms generated from EEG signals. The SVM achieved an accuracy of 71.6%, 74.76%, 70.70% and 81.47% using GoogleNet, Mobilenet, AlexNet and SqueezeNet for scalogram feature extraction. The RVM achieved an accuracy of 65.5%, 69.9%, 65.3% and 72.59% when used for classification using the features generated from GoogLeNet, AlexNet, MobileNet and SqueezeNet.The SqueezeNet deep CNN performed better than GoogLeNet, AlexNet and MobileNet for classification of the EEG scalograms. The feature extraction using SqueezeNet also resulted in better classification accuracy obtained by SVM and RVM. The results indicate that pre-trained models can be used for classifying the ASD using scalograms of the EEG signals.

EMD-WOG-2DCNN based EEG signal processing for Rolandic seizure classification

Objective Approximately 65 million people have epilepsy around the world. Recognition of epilepsy types is the basis to determine the treatment method and predict the prognosis in epilepsy patients. Childhood benign epilepsy with centrotemporal spikes (BECTS) or benign Rolandic epilepsy is the most common focal epilepsy in children, accounting for 15–20% of childhood epilepsies. These EEG patterns of individuals usually predict good treatment responses and prognosis. Until now, the interpretation of EEG still depends entirely on experienced neurologists, which may be a lengthy and tedious task. Method In this article, we proposed a novel machine learning model that efficiently distinguished Rolandic seizures from normal EEG signals. The proposed machine learning model processes the identification procedure in the following order (1) creating preliminary EEG features using signal empirical mode decomposition, (2) applying weighted overlook graph (WOG) to represent the decomposed EMD of IMF, and (3) classifying the results through a two Dimensional Convolutional Neural Network (2DCNN). The performance of our classification model is compared with other representative machine learning models. Results The model offered in this article gains an accuracy performance exceeding 97.6% in the Rolandic dataset, which is higher than other classification models. The effect of the model on the Bonn public dataset is also comparable to existing methods and even performs better in some subsets. Conclusion The purpose of this study is to introduce the most common childhood benign epilepsy type and propose a model that meets the real clinical needs to distinguish this Rolandic EEG pattern from normal signals accurately. Significance Future research will optimize the model to categorize other types of epilepsies beyond BECTS and finally implement them in the hospital system.

Semi-supervised automatic seizure detection using personalized anomaly detecting variational autoencoder with behind-the-ear EEG

Background and ObjectiveEpilepsy is one of the most common neurologic diseases worldwide, and 30% of the patients live with uncontrolled seizures. For the safety of patients with epilepsy, an automatic seizure detection algorithm for continuous seizure monitoring in daily life is important to reduce risks related to seizures, including sudden unexpected death. Previous researchers applied machine learning to detect seizures with EEG, but the epileptic EEG waveform contains subtle changes that are difficult to identify. Furthermore, the imbalance problem due to the small proportion of ictal events caused poor prediction performance in supervised learning approaches. This study aimed to present a personalized deep learning-based anomaly detection algorithm for seizure monitoring with behind-the-ear electroencephalogram (EEG) signals. MethodsWe collected behind-the-ear EEG signals from 16 patients with epilepsy in the hospital and used them to develop and evaluate seizure detection algorithms. We modified the variational autoencoder network to learn the latent representation of normal EEG signals and performed seizure detection by measuring the anomalies in EEG signals using the trained network. To personalize the algorithm, we also proposed a method to calibrate the anomaly score for each patient by comparing the representations in the latent space. ResultsOur proposed algorithm showed a sensitivity of 90.4% with a false alarm rate of 0.83 per hour without personal calibration. On the other hand, the one-class support vector machine only showed a sensitivity of 84.6% with a false alarm rate of 2.17 per hour. Furthermore, our proposed model with personal calibration achieved 94.2% sensitivity with a false alarm rate of 0.29 while detecting 49 of 52 ictal events. ConclusionsWe proposed a novel seizure detection algorithm with behind-the-ear EEG signals via semi-supervised learning of an anomaly detecting variational autoencoder and personalization method of anomaly scoring by comparing latent representations. Our approach achieved improved seizure detection with high sensitivity and a lower false alarm rate.

Detection of normal and epileptic EEG signals using by lifting based HAAR wavelet transform and artificial neural network

Detection of normal and epileptic EEG signals using by lifting based HAAR wavelet transform and artificial neural network