Narrow Results

This paper presents a cognitive state estimation system focused on some effective feature extraction based on temporal and spectral analysis of electroencephalogram (EEG) signal and the proper channel selection of the BIOPAC automated EEG analysis system. In the proposed approach, different frequency components (i) real value; (ii) imaginary value; (iii) magnitude; (iv) phase angle and (v) power spectral density of the EEG data samples during different mental task performed to assess seven types of human cognitive states — relax, mental task, memory related task, motor action, pleasant, fear and enjoying music on the three channels of BIOPAC EEG data acquisition system — EEG, Alpha and Alpha RMS signal. Also the time and time-frequency-based features were extracted to compare the performance of the system. After feature extraction, the channel efficacy is evaluated by support vector machine (SVM) based on the classification rate in different cognitive states. From the experimental results and classification accuracy, it is determined that the overall accuracy for alpha channel shows much improved result for power spectral density than the other frequency based features and other channels. The classification rate is 69.17% for alpha channel whereas for EEG and alpha RMS channel it is found 47.22% and 32.21%, respectively. For statistical analysis standard deviation shows better result for alpha channel and it is found 65.4%. The time-frequency analysis shows much improved result for alpha channel also. For the mean value of DWT coefficients the accuracy is highest and it is 81.3%. Besides the classification accuracy, SVM shows better performance in compare with kNN classifier.

This paper presents a method to characterize, identify and classify some pathological Electroencephalogram (EEG) signals. We use some Time Frequency Distributions (TFDs) to analyze its nonstationarity. The analysis is conducted by the spectrogram (SP), the Choi–Williams Distribution (CWD) and the Smoothed Pseudo Wigner Ville Distribution (SPWVD). The studies are carried on some real EEG signals collected from a known database. The estimation of the best value of parameters for each distribution is achieved using the Rényi entropy (RE). The time-frequency results have permitted to characterize some pathological EEG signals. In addition, the Rényi Marginal Entropy (RME) is used for the purpose of detecting the peak seizures and discriminates between normal and pathological EEG signals. The frequency bands are evaluated using the Marginal Frequency (MF). The EEG signal classification of two sets A and E containing normal and pathologic EEG signals, respectively, is performed using our proposed method based on energy extraction of signals from time-frequency plane. Also, the Moving Average (MA) is used as a tool to obtain better classification results. The results conducted on real-life EEG signals illustrate the effectiveness of the proposed method.

This paper presents a fully integrated front-end, low noise amplifier (LNA), dedicated to the processing of various types of bio-medical signals, such as Electrocardiogram (ECG), Electroencephalography (EEG), Axon Action Potential (AAP). A novel noise reduction technique, for an operational transconductance amplifier (OTA), has been proposed. This adds a current steering branch parallel to the differential pair, with a view to reducing the noise contribution by the cascode current sources. Hence, this reduces the overall input-referred noise of the LNA, without adding any additional power. The proposed technique implemented in 65nm CMOS technology achieves 30 dB closed-loop voltage gain, 0.05Hz lower cut-off frequency and 100 MHz 3-dB bandwidth. It operates at 1.2V power supply and draws 1$\mu$A static current. The prototype described in this paper occupies 3300$\mu$m² silicon area.

One of the important parameters in the brain–computer interface (BCI) system is speed. Therefore, it is always desirable to design a high-speed system that has an acceptable performance, simultaneously. The main idea of this paper is the use of evolutionary algorithms (EAs) to select the optimal features for epilepsy diagnosis by processing the electroencephalogram (EEG) signals. The lesser the number of features is, the higher will be the usefulness of accuracy of the system to us. Therefore, here, using EAs, some of the features that are redundant in the data and do not contain a lot of information and only increase the complexity of the system are eliminated and the best features are chosen. We select this choice by EAs. Running the feature selection step is after the feature extraction step. In fact, the features were extracted using the common spatial pattern (CSP) algorithm, and then the optimal features were selected from the extracted feature set. This can save a lot of system complexity and reduce system execution time considerably. Finally, at the diagnostic stage, these selected features are given to a simple neural network (NN). The results showed that when the combination of EA and CSP is used, the precision of the system is much higher than when the CSP method is only used, although it contributes significantly to the complexity of the system.

Inverse algorithms are used to assess EEG source parameters. This involves identifying unknown voxels in hundreds of different regions, giving an incomplete picture of the brain. There are no uniform solutions since the same sensor output may come from many source configurations. To overcome the lack of uniqueness, one must consider previous information and parameters inherent in the source. Our goal is to predict the location of brain sources from the recorded EEG signal without any prior knowledge of sources. We applied a particle filter to locate the brain sources in this article. The degeneracy of particle weights limits the particle filter’s performance. Various resampling techniques are suggested to address this problem. The performance of the branching resampling approach is compared to a systematic resampling method for brain source localization. To perform assessment and comparison studies, both simulated and real EEG data are used.

EEG-based brain–computer interface systems have demonstrated impressive capability in the context of the broad use of deep learning. However, serious security flaws have been shown by their vulnerability to adversarial sample attacks. It is not possible to directly apply traditional attack strategies from the image domain to EEG signals because of their unique visualization and dynamic features. A unique multi-target smooth universal adversarial perturbation (MS-UAP) approach designed specifically for EEG data is presented in this paper to address this difficulty. We define a loss function and optimize parameters based on UAP for both target and nontarget class samples, respectively, to address the temporal nature of the EEG signal and the attack’s flexibility. Additionally, this research generates a smoothed adversarial sample by convolution operations, yielding a more disorienting effect, to avoid the introduction of physiologically untrustworthy square wave distortions caused by typical visual attack approaches. Three public EEG datasets were used for extensive testing and assessments, and the results show that MS-UAP successfully targets convolutional neural network (CNN) classifiers and launches simultaneous attacks on several classes, exhibiting strong model transferability. To ensure accurate attacks that go undetected, MS-UAP creates a single disturbance that is targeted at target classes and has the least negative effect on other classes. In order to facilitate effective real-time attacks, MS-UAP, an offline-generated universal perturbation template, can be smoothly incorporated into EEG data. The study is significant since it emphasizes how important it is to improve BCI system security and promote the creation of stronger defenses.

Many studies developed the machine learning method for discriminating Major Depressive Disorder (MDD) and normal control based on multi-channel electroencephalogram (EEG) data, less concerned about using single channel EEG collected from forehead scalp to discriminate the MDD. The EEG dataset is collected by the Fp1 and Fp2 electrode of a 32-channel EEG system. The result demonstrates that the classification performance based on the EEG of Fp1 location exceeds the performance based on the EEG of Fp2 location, and shows that single-channel EEG analysis can provide discrimination of MDD at the level of multi-channel EEG analysis. Furthermore, a portable EEG device collecting the signal from Fp1 location is used to collect the second dataset. The Classification and Regression Tree combining genetic algorithm (GA) achieves the highest accuracy of 86.67% based on leave-one-participant-out cross validation, which shows that the single-channel EEG-based machine learning method is promising to support MDD prescreening application.

An analysis of the Electroencephalogram (EEG) signals while performing a monotonous task and drinking alcohol using principal component analysis (PCA), linear discriminant analysis (LDA) for feature extraction and Neural Networks (NNs) for classification is proposed. The EEG is captured while performing a monotonous task that can adversely affect the brain and possibly cause stress. Moreover, we investigate the effects of alcohol on the brain by capturing the data continuously after consumption of equal amounts of alcohol. We hope that our work will shed more light on the relationship between such actions and EEG, and investigate if there is any relation between the tasks and mental stress. EEG signals offers a rare look at brain activity, while, monotonous activities are well known to cause irritation which may contribute to mental stress. We apply PCA and LDA to characterize the change in each component, extract it and discriminate using a NN. After experiments, it was found that PCA and LDA are effective analysis methods in EEG signal analysis.

Epilepsy is referred to as a neurological disorder, which is detected via examination and manual comprehension of Electroencephalogram (EEG) signals. In deep learning schemes, various enhancements have emerged to efficiently address complex issues by end-to-end learning. The major objective of this research is to propose a new seizure detection approach from EEG signals using a deep learning-based classification technique. The pre-processing is the initial stage, where denoising is performed using a Short-Time Fourier Transform (STFT). Subsequently, the statistical features, time-domain features and spectral features are extracted from the pre-processed signal. Finally, an efficient optimization approach, named Adadelta-Chameleon Swarm Algorithm (Adadelta-CSA), is proposed and employed to train Deep Neural Network (DNN) to carry out the precise seizure prediction. Here, the integration of the Adadelta concept in the Chameleon Swarm Algorithm (CSA) has resulted in Adadelta-CSA. At last, the performance of the Adadelta-CSA scheme-based DNN is compared with the existing techniques by considering accuracy, sensitivity and specificity, and it is found to produce better values of 0.951, 0.966, and 0.935, respectively.

The analysis of biological fluctuations provides an excellent route to probe the underlying mechanisms in maintaining internal homeostasis of the body, especially under the challenges of the ever-changing environment or disease processes. However, the features of nonlinearity and nonstationarity in physiological time series limit the reliability of the conventional analysis. Hilbert–Huang transform (HHT), based on nonlinear theory, is an innovative approach to extract the dynamic information at different time scales, in particular, from nonstationary signals. In this paper, HHT is introduced to analyze the alpha waves of human's electroencephalography (EEG), which seemly oscillate regularly between 8 and 12 Hz in healthy subject but getting irregular or disappeared in different demented status. Furthermore, conventional time–frequency analyses are adopted to collate the results from those methods and HHT. Finally, the potential usages of HHT are demonstrated in characterizing the biological signals qualitatively and quantitatively, including stationarity analysis, instantaneous frequency and amplitude modulation or correlation analysis. Such applications on EEG have successively disclosed the differences of alpha rhythms between normal and demented brains and the nonlinear characteristics of the underlying mechanisms. Hopefully, in addition to empower the studies of EEG varied in diseased, aging, and physiological processes, these methods might find other applications in EEG analysis.

Empirical mode decomposition (EMD) provides an adaptive, data-driven approach to time–frequency analysis, yielding components from which local amplitude, phase, and frequency content can be derived. Since its initial introduction to electroencephalographic (EEG) data analysis, EMD has been extended to enable phase synchrony analysis and multivariate data processing. EMD has been integrated into a wide range of applications, with emphasis on denoising and classification. We review the methodological developments, providing an overview of the diverse implementations, ranging from artifact removal to seizure detection and brain–computer interfaces. Finally, we discuss limitations, challenges, and opportunities associated with EMD for EEG analysis.

This study is to examine the correlation between customers’ interactive experience with flower stores and their purchase intentions through the aesthetic experience along with the examination of customers’ brainwaves. Among a total of 56 effective questionnaires, 10 random participants are chosen to give electroencephalogram (EEG) examinations and interviews. Regression analyses showed the explained variance 57.9% and the regression coefficients 0.415 (interactive experience) and 0.422 (esthetic experience). The EEG in interactive experiences varied case by case, but delta tended to increase in aesthetic experiences.

Utilizing electroencephalography (EEG) for emotion recognition enables the direct collection of physiological signals, thereby circumventing potential deception associated with facial expressions and language cues. Concurrently, there has been a widespread adoption of deep learning techniques in this domain, aimed at achieving enhanced results through model and parameter adjustments. An innovative approach is introduced by incorporating the Fast Fourier Transform (FFT) for data preprocessing in conjunction with the CBi-GRU model, merging the Convolutional Neural Network (CNN) and the Bidirectional Gated Recurrent Unit (BiGRU). Comparative analysis underscores the significance of exploring the intrinsic characteristics of the data itself: the application of FFT not only substantially reduces processing time but also enhances model performance. The integration of CBiGRU and FFT leverages the strengths of both CNN and BiGRU methods, resulting in increased accuracy and rapid convergence.

The human brain has several major levels of functioning, e.g.: quantum, genetic, single neuron, neural networks, cognitive, each of them involving a complex dynamic interaction between participating elements and signals. There are also complex dynamic interactions across functional levels [1].

With the advancement of the bioinformatics and brain research technologies more data become available tracing the activity of genes, neurons, neural networks and brain areas over time [2,3,4]. How can such data be used to create a dynamic model that captures dynamic interactions at a particular level over time? How can we integrate dynamic models at different levels? How do we use these models to better understand brain dynamics? These are the main questions addressed in the paper through dynamic interaction network modelling.

Dynamic Interaction Networks (DIN) are popular methods [1] especially for modelling a biological gene regulatory network (GRN) of n genes expressed over time. A node Nj(t) in the model represents the gene Gj(t) activation (expression) at a particular time t and the weighted arcs Wij represent the degree of interaction between genes Gi (i=1,2,…,n) and Gj at two consecutive time moments t and (t+1). In order to evaluate Nj(t+1) a function Fj (Ni, Wij; i=1,2,…,n) is used. The regulatory network model is created through optimisation procedure that optimises the connection weight matrix W, the functions Fj (j=1,2,…,n) from time course gene expression data. There are several methods used so far to create a GRN from time course data.

The paper demonstrates how DIN can be used to model brain dynamics at different levels. At a genetic level, a GRN can be created for a single neuron involving a set of relevant to a particular neuronal function genes. At a next, more complex level, a GRN can be created to capture the dynamics of the expression of several genes related to a complex brain function of a brain area, such as LTP in the CA3 area of the hippocampus. A next level of complexity is to create a model that includes a GRN and matches not only gene expression values over time, but brain signals as well, such as EEG, related to a particular brain function or a disease, e.g. epilepsy. All these models are called computational neurogenetic models [5].

A DIN can also be created from a time series of EEG measurements related to perceptual or cognitive functions, such as visual and auditory perception. The DIN model represents EEG channels as nodes and the connection weight matrix represents the temporal interaction between the brain signals measured by the channels. It is demonstrated that building a DIN model from EEG data can help tracing and discovering hidden brain signal interactions for particular brain functions. Comparing DIN models, built on EEG data measured for different stimuli and different individuals, can also help understand differences in the brain functioning for different perception and cognitive tasks and for different individuals.

The paper demonstrates all the above levels of DIN modelling and illustrates them on real brain data. The paper suggests directions for further research in modelling and discovery of dynamic brain interactions relating to complex brain functions at different functional level.

There is strong evidence pointing to the existence of sudden jumps and apparent discontinuities in spatio-temporal correlates of cognitive activity over the theta frequency band. This topic, however, is highly controversial as it requires revision of our traditional view on brain functions and the nature of intelligence. Human cognition performs a granulation of the seemingly homogeneous temporal sequences of perceptual experiences into meaningful and comprehendible chunks of fuzzy concepts and complex behavioral schemas. They are accessed during future action selection and decisions. In this work a dynamical approach to higher cognition and intelligence is presented to interpret experimental findings. In the model, meaningful knowledge is continuously created, processed, and dissipated in the form of sequences of oscillatory patterns of neural activity distributed across space and time. Oscillatory patterns can be viewed as intermittent manifestations of a generalized symbol system, with which brains compute. These patterns are not rigid but flexible and they dissipate soon after they have been generated through spatio-temporal phase transitions. The proposed biologically-motivated computing using dynamic patterns provides an alternative to the notoriously difficult symbol grounding problem and it has been implemented in computational and robotic environments.

Finding human cognitive state is very important and useful for medical cares in our daily life. Eye state classification is a kind of common time-series problem for detecting human cognitive state. One of approaches of EEG eye state classification is based on time series data. We propose a new neural fuzzy structure that is possible to use an eye state classification. The classification accuracy can be achieved by using the proposed approach in terms of the number of neurons in the hidden layer, which also leads types of membership function in fuzzy rules. The data of EEG signals for monitoring eye state saved by Machine Learning Repository, University of California, Irvine (UCI) is used for eye state benchmarking. We did experiments with 3 different networks for the architecture by changing the number of neurons in the hidden layer and random seed for weights. We found that tuning parameters asymmetrically gave us the best results through test cases. According to the test results, we have the best result with 4 neurons by managing parameters of standard deviations asymmetrically, with which showed a 4.0% average error rate with the test data.