RESUMO
We present a new approach to classifying the sleep stage that incorporates a computationally inexpensive method based on permutations for channel selection and takes advantage of deep learning power, specifically the gated recurrent unit (GRU) model, along with other deep learning methods. By systematically permuting the electroencephalographic (EEG) channels, different combinations of EEG channels are evaluated to identify the most informative subset for the classification of the 5-class sleep stage. For analysis, we used an EEG dataset that was collected at the International Institute for Integrative Sleep Medicine (WPI-IIIS) at the University of Tsukuba in Japan. The results of these explorations provide many new insights such as the (1) drastic decrease in performance when channels are fewer than 3, (2) 3-random channels selected by permutation provide the same or better prediction than the 3 channels recommended by the American Academy of Sleep Medicine (AASM), (3) N1 class suffers the most in prediction accuracy as the channels drop from 128 to 3 random or 3 AASM, and (4) no single channel provides acceptable levels of accuracy in the prediction of 5 classes. The results obtained show the GRU's ability to retain essential temporal information from EEG data, which allows capturing the underlying patterns associated with each sleep stage effectively. Using permutation-based channel selection, we enhance or at least maintain as high model efficiency as when using high-density EEG, incorporating only the most informative EEG channels.
Assuntos
Eletroencefalografia , Fases do Sono , Humanos , Eletroencefalografia/métodos , Fases do Sono/fisiologia , Aprendizado Profundo , Masculino , Feminino , Adulto , Polissonografia/métodosRESUMO
High-density Electroencephalography (HD-EEG) has proven to be the EEG montage that estimates the neural activity inside the brain with highest accuracy. Multiple studies have reported the effect of electrode number on source localization for specific sources and specific electrode configurations. The electrodes for these configurations are often manually selected to uniformly cover the entire head, going from 32 to 128 electrodes, but electrode configurations are not often selected according to their contribution to estimation accuracy. In this work, an optimization-based study is proposed to determine the minimum number of electrodes that can be used and to identify the optimal combinations of electrodes that can retain the localization accuracy of HD-EEG reconstructions. This optimization approach incorporates scalp landmark positions of widely used EEG montages. In this way, a systematic search for the minimum electrode subset is performed for single- and multiple-source localization problems. The Non-dominated Sorting Genetic Algorithm II (NSGA-II) combined with source reconstruction methods is used to formulate a multi-objective optimization problem that concurrently minimizes (1) the localization error for each source and (2) the number of required EEG electrodes. The method can be used for evaluating the source localization quality of low-density EEG systems (e.g. consumer-grade wearable EEG). We performed an evaluation over synthetic and real EEG datasets with known ground-truth. The experimental results show that optimal subsets with 6 electrodes can attain an equal or better accuracy than HD-EEG (with more than 200 channels) for a single source case. This happened when reconstructing a particular brain activity in more than 88% of the cases in synthetic signals and 63% in real signals, and in more than 88% and 73% of cases when considering optimal combinations with 8 channels. For a multiple-source case of three sources (only with synthetic signals), it was found that optimized combinations of 8, 12 and 16 electrodes attained an equal or better accuracy than HD-EEG with 231 electrodes in at least 58%, 76%, and 82% of cases respectively. Additionally, for such electrode numbers, lower mean errors and standard deviations than with 231 electrodes were obtained.
Assuntos
Mapeamento Encefálico , Eletroencefalografia , Algoritmos , Mapeamento Encefálico/métodos , Eletrodos , Eletroencefalografia/métodos , Couro CabeludoRESUMO
In this study we explore how different levels of emotional intensity (Arousal) and pleasantness (Valence) are reflected in electroencephalographic (EEG) signals. We performed the experiments on EEG data of 32 subjects from the DEAP public dataset, where the subjects were stimulated using 60-s videos to elicitate different levels of Arousal/Valence and then self-reported the rating from 1 to 9 using the self-assessment Manikin (SAM). The EEG data was pre-processed and used as input to a convolutional neural network (CNN). First, the 32 EEG channels were used to compute the maximum accuracy level obtainable for each subject as well as for creating a single model using data from all the subjects. The experiment was repeated using one channel at a time, to see if specific channels contain more information to discriminate between low vs high arousal/valence. The results indicate than using one channel the accuracy is lower compared to using all the 32 channels. An optimization process for EEG channel selection is then designed with the Non-dominated Sorting Genetic Algorithm II (NSGA-II) with the objective to obtain optimal channel combinations with high accuracy recognition. The genetic algorithm evaluates all possible combinations using a chromosome representation for all the 32 channels, and the EEG data from each chromosome in the different populations are tested iteratively solving two unconstrained objectives; to maximize classification accuracy and to reduce the number of required EEG channels for the classification process. Best combinations obtained from a Pareto-front suggests that as few as 8-10 channels can fulfill this condition and provide the basis for a lighter design of EEG systems for emotion recognition. In the best case, the results show accuracies of up to 1.00 for low vs high arousal using eight EEG channels, and 1.00 for low vs high valence using only two EEG channels. These results are encouraging for research and healthcare applications that will require automatic emotion recognition with wearable EEG.
Assuntos
Eletroencefalografia , Redes Neurais de Computação , Algoritmos , Nível de Alerta , Eletroencefalografia/métodos , Emoções , HumanosRESUMO
We present a new approach for a biometric system based on electroencephalographic (EEG) signals of resting-state, that can identify a subject and reject intruders with a minimal subset of EEG channels. To select features, we first use the discrete wavelet transform (DWT) or empirical mode decomposition (EMD) to decompose the EEG signals into a set of sub-bands, for which we compute the instantaneous and Teager energy and the Higuchi and Petrosian fractal dimensions for each sub-band. The obtained features are used as input for the local outlier factor (LOF) algorithm to create a model for each subject, with the aim of learning from it and rejecting instances not related to the subject in the model. In search of a minimal subset of EEG channels, we used a channel-selection method based on the non-dominated sorting genetic algorithm (NSGA)-III, designed with the objectives of minimizing the required number EEG channels and increasing the true acceptance rate (TAR) and true rejection rate (TRR). This method was tested on EEG signals from 109 subjects of the public motor movement/imagery dataset (EEGMMIDB) using the resting-state with the eyes-open and the resting-state with the eyes-closed. We were able to obtain a TAR of [Formula: see text] and TRR of [Formula: see text] using 64 EEG channels. More importantly, with only three channels, we were able to obtain a TAR of up to [Formula: see text] and a TRR of up to [Formula: see text] for the Pareto-front, using NSGA-III and DWT-based features in the resting-state with the eyes-open. In the resting-state with the eyes-closed, the TAR was [Formula: see text] and the TRR [Formula: see text] also using DWT-based features from three channels. These results show that our approach makes it possible to create a model for each subject using EEG signals from a reduced number of channels and reject most instances of the other 108 subjects, who are intruders in the model of the subject under evaluation. Furthermore, the candidates obtained throughout the optimization process of NSGA-III showed that it is possible to obtain TARs and TRRs above 0.900 using LOF and DWT- or EMD-based features with only one to three EEG channels, opening the way to testing this approach on bigger datasets to develop a more realistic and usable EEG-based biometric system.
Assuntos
Algoritmos , Biometria/métodos , Encéfalo/fisiologia , Interpretação Estatística de Dados , Eletroencefalografia/instrumentação , Eletroencefalografia/métodos , Humanos , Processamento de Sinais Assistido por Computador , Análise de OndaletasRESUMO
We present a multi-objective optimization method for electroencephalographic (EEG) channel selection based on the non-dominated sorting genetic algorithm (NSGA) for epileptic-seizure classification. We tested the method on EEG data of 24 patients from the CHB-MIT public dataset. The procedure starts by decomposing the EEG data from each channel into different frequency bands using the empirical mode decomposition (EMD) or the discrete wavelet transform (DWT), and then for each sub-band four features are extracted; two energy values and two fractal dimension values. The obtained feature vectors are then iteratively tested for solving two unconstrained objectives by NSGA-II or NSGA-III; to maximize classification accuracy and to reduce the number of EEG channels required for epileptic seizure classification. Our results have shown accuracies of up to 1.00 with only one EEG channel. Interestingly, when using all the EEG channels available, lower accuracies were achieved compared to the case when EEG channels were selected by NSGA-II or NSGA-III; i.e., in patient 19 we obtained an accuracy of 0.95 using all the channels and 0.975 using only two channels selected by NSGA-III. The results obtained are encouraging and it has been shown that it is possible to classify epileptic seizures using a few electrodes, which provide evidence for the future development of portable EEG seizure detection devices.
RESUMO
We present a four-objective optimization method for optimal electroencephalographic (EEG) channel selection to provide access to subjects with permission in a system by detecting intruders and identifying the subject. Each instance was represented by four features computed from two sub-bands, extracted using empirical mode decomposition (EMD) for each channel, and the feature vectors were used as input for one-class/multi-class support vector machines (SVMs). We tested the method on data from the event-related potentials (ERPs) of 26 subjects and 56 channels. The optimization process was performed by the non-dominated sorting genetic algorithm (NSGA), which found a three-channel combination that achieved an accuracy of 0.83, with both a true acceptance rate (TAR) and a true rejection rate (TRR) of 1.00. In the best case, we obtained an accuracy of up to 0.98 for subject identification with a TAR of 0.95 and a TRR 0.93, all using seven EEG channels found by NSGA-III in a subset of subjects manually created. The findings were also validated using 10 different subdivisions of subjects randomly created, obtaining up to 0.97 ± 0.02 of accuracy, a TAR of 0.81 ± 0.12 and TRR of 0.85 ± 0.10 using eight channels found by NSGA-III. These results support further studies on larger datasets for potential applications of EEG in identification and authentication systems.
Assuntos
Eletroencefalografia/métodos , Sistemas de Identificação de Pacientes/métodos , Adulto , Algoritmos , Encéfalo/fisiologia , Potenciais Evocados/fisiologia , Humanos , Reprodutibilidade dos Testes , Máquina de Vetores de SuporteRESUMO
We are here to present a new method for the classification of epileptic seizures from electroencephalogram (EEG) signals. It consists of applying empirical mode decomposition (EMD) to extract the most relevant intrinsic mode functions (IMFs) and subsequent computation of the Teager and instantaneous energy, Higuchi and Petrosian fractal dimension, and detrended fluctuation analysis (DFA) for each IMF. We validated the method using a public dataset of 24 subjects with EEG signals from 22 channels and showed that it is possible to classify the epileptic seizures, even with segments of six seconds and a smaller number of channels ( e.g., an accuracy of 0.93 using five channels). We were able to create a general machine-learning-based model to detect epileptic seizures of new subjects using epileptic-seizure data from various subjects, after reducing the number of instances, based on the k-means algorithm.