Cognitive Neurodynamics最新文献

The electroencephalogram (EEG) signal is being investigated as a more confidential biometric for person identification. Despite recent advancements, a persistent challenge lies in the influence of variations in affective states. Affective states consistently exist during data collection, regardless of the protocol used. Additionally, the inherently non-stationary nature of EEG makes it susceptible to fluctuations in affective states over time. Therefore, it would be highly crucial to perform precise EEG-based person identification under varying affective states. This paper employed an integrated Multi-scale Convolution and Graph Pooling network (MCGP) to mitigate the impact of affective state variations. MCGP utilized multiple 1D convolutions at different scales to dynamically extract and fuse features. Additionally, a graph pooling layer with an attention mechanism was incorporated to generate hierarchical graph embeddings. These embeddings were concatenated as inputs for a fully connected classification layer. Experiments were conducted on the SEED and SEED-V dataset, revealing that MCGP achieved an average accuracy of 85.51% for SEED and 88.69% for SEED-V in cross-session conditions involving mixed affective states. Under single affective state cross-session scenario, MCGP achieved an average accuracy of 85.75% for SEED and 88.06% for SEED-V for the same affective states, while obtaining 79.57% for SEED and 84.52% for SEED-V for different affective states. Results indicated that, compared to the baseline methods, MCGP effectively mitigated the impact of variations in affective states across different sessions. In single affective state cross-session scenario, identification performance for the same affective states was slightly higher than that for different affective states.

Emotion recognition plays an important role in human life and healthcare. The EEG has been extensively researched as an objective indicator of intense emotions. However, current existing methods lack sufficient analysis of shallow and deep EEG features. In addition, human emotions are complex and variable, making it difficult to comprehensively represent emotions using a single-modal signal. As a signal associated with gaze tracking and eye movement detection, Eye-related signals provide various forms of supplementary information for multimodal emotion analysis. Therefore, we propose a Pseudo-Siamese Pyramid Network (PSPN) for multimodal emotion analysis. The PSPN model employs a Depthwise Separable Convolutional Pyramid (DSCP) to extract and integrate intrinsic emotional features at various levels and scales from EEG signals. Simultaneously, we utilize a fully connected subnetwork to extract the external emotional features from eye-related signals. Finally, we introduce a Pseudo-Siamese network that integrates a flexible cross-modal dual-branch subnetwork to collaboratively utilize EEG emotional features and eye-related behavioral features, achieving consistency and complementarity in multimodal emotion recognition. For evaluation, we conducted experiments on the DEAP and SEED-IV public datasets. The experimental results demonstrate that multimodal fusion significantly improves the accuracy of emotion recognition compared to single-modal approaches. Our PSPN model achieved the best accuracy of 96.02% and 96.45% on the valence and arousal dimensions of the DEAP dataset, and 77.81% on the SEED-IV dataset, respectively. Our code link is: https://github.com/Yinyanyan003/PSPN.git.

In brain-computer interfaces (BCIs) based on motor imagery (MI), reducing calibration time is gradually becoming an urgent issue in practical applications. Recently, transfer learning (TL) has demonstrated its effectiveness in reducing calibration time in MI-BCI. However, the different data distribution of subjects greatly affects the application effect of TL in MI-BCI. Therefore, this paper combines data alignment, source domain selection, and feature alignment into the MI-TL. We propose a novel dual-step transfer framework based on source domain selection and feature alignment. First, the source and target domains are aligned using a pre-calibration strategy (PS), and then a sequential reverse selection method is proposed to match the optimal source domain for each target domain with the designed dual model selection strategy. We use filter bank regularization common space pattern (FBRCSP) to obtain more features and introduce manifold embedded distribution alignment (MEDA) to correct the prediction results of the support vector machine (SVM). The experimental results on two competition public datasets (BCI competition IV Dataset 1 and Dataset 2a) and our dataset show that the average classification accuracy of the proposed framework is higher than the baseline method (no domain selection and no feature alignment), which reaches 84.12%, 79.91%, and 78.45%, respectively. And the computational cost is reduced by half compared with the baseline method.

Brain-computer interface (BCI)-based robot combines BCI and robotics technology to realize the brain’s intention to control the robot, which not only opens up a new way for the daily care of the disabled individuals, but also provides a new way of communication for normal people. However, the existing systems still have shortcomings in many aspects such as friendliness of human–computer interaction, and interaction efficient. This study developed a humanoid robot control system by integrating an augmented reality (AR)-based BCI with a simultaneous localization and mapping (SLAM)-based scheme for autonomous indoor navigation. An 8-target steady-state visual evoked potential (SSVEP)-based BCI was implemented to enable direct control of the humanoid robot by the user. A Microsoft HoloLens was utilized to display visual stimuli for eliciting SSVEPs. Filter bank canonical correlation analysis (FBCCA), a training-free method, was used to detect SSVEPs in this study. By leveraging SLAM technology, the proposed system alleviates the need for frequent control commands transmission from the user, thereby effectively reducing their workload. Online results from 12 healthy subjects showed this developed BCI system was able to select a command out of eight potential targets with an average accuracy of 94.79%. The autonomous navigation subsystem enabled the humanoid robot to autonomously navigate to a destination chosen utilizing the proposed BCI. Furthermore, all participants successfully completed the experimental task using the developed system without any prior training. These findings illustrate the feasibility of the developed system and its potential to contribute novel insights into humanoid robots control strategies.

This research explores the various chaotic features of the hyperchaotic Chen dynamical system within a variable order fractional (VOF) calculus framework, employing an innovative approach with a nonlinear and adaptive radial basis function neural network. The study begins by computing the numerical solution of VOF differential equations for the hyperchaotic Chen system through a numerical scheme using the Caputo–Fabrizio derivative across a spectrum of different system control parameters. Subsequently, a comprehensive parametric model is formulated using RBFNN, considering the system’s various initial values. We systematically investigate the various chaotic attractors of the proposed system, employing statistical analysis, phase space reconstruction, and Lyapunov exponent. Additionally, we assess the effectiveness of the proposed computational RBFNN model using the Root Mean Square Error statistic. Importantly, the obtained results closely align with those derived from numerical algorithms, emphasizing the high accuracy and reliability of the designed network. The outcomes of this study have implications for studying chaos with variable fractional derivatives, with applications across various scientific and engineering domains. This work advances the understanding and applications of variable order fractional dynamics.

The basal ganglia (BG) show a variety of functions for motor and cognition. There are two competitive pathways in the BG; direct pathway (DP) which facilitates movement and indirect pathway (IP) which suppresses movement. It is well known that diverse functions of the BG may be made through “balance” between DP and IP. But, to the best of our knowledge, so far no quantitative analysis for such balance was done. In this paper, as a first time, we introduce the competition degree ({{mathcal {C}}}_d) between DP and IP. Then, by employing ({{mathcal {C}}}_d), we quantify their competitive harmony (i.e., competition and cooperative interplay), which could lead to improving our understanding of the traditional “balance” so clearly and quantitatively. We first consider the case of normal dopamine (DA) level of (phi ^*=0.3). In the case of phasic cortical input (10 Hz), a healthy state with ({{mathcal {C}}}_d^* = 2.82) (i.e., DP is 2.82 times stronger than IP) appears. In this case, normal movement occurs via harmony between DP and IP. Next, we consider the case of decreased DA level, (phi = phi ^*(=0.3)~x_{DA}) ((1 > x_{DA} ge 0)). With decreasing (x_{DA}) from 1, the competition degree ({{mathcal {C}}}_d) between DP and IP decreases monotonically from ({{mathcal {C}}}_d^*), which results in appearance of a pathological Parkinsonian state with reduced ({{mathcal {C}}}_d). In this Parkinsonian state, strength of IP is much increased than that in the case of normal healthy state, leading to disharmony between DP and IP. Due to such break-up of harmony between DP and IP, impaired movement occurs. Finally, we also study treatment of the pathological Parkinsonian state via recovery of harmony between DP and IP.

Numerous studies on early detection of schizophrenia (SZ) have utilized all available channels or employed set of a few time domain or frequency domain features, while a limited number of features may not be sufficient enough to perform diagnosis efficiently. To encounter these problems, an automated diagnosis model is proposed for the efficient diagnosis of schizophrenia symptomatic adolescent subjects from electroencephalogram (EEG) signals via machine intelligence. A publicly accessible EEG dataset featuring 16-channels EEG obtained from 84 adolescents (45 SZ symptomatic and 39 healthy control) is used to demonstrate the work. Initially, the signals are decomposed into sub-bands using two multi-resolution signal analysis methods: Empirical Wavelet Transform and Empirical mode decomposition. 75 unique features from each sub-bands are extracted and the few selective prominent features are applied to machine learning classifiers for optimal sub-band selection. Subsequently, a hybrid model is proposed, combining convolutional neural network (CNN) and ensemble bagged tree, incorporating both deep learning and handcrafted features to perform SZ diagnosis. This innovative model achieved superior classification performance compared to existing methods, offering a promising approach for SZ diagnosis. Furthermore, the study explores the impact of different brain regions and combined regional data in SZ diagnosis comprehensively. Hence, this computer-assisted decision-making model minimizes the limitations of prior studies by providing a more robust and efficient diagnostic system for schizophrenia.

Chaos is often described as the limited development of nonlinear dynamic systems that create intricate and non-repetitive patterns. In this study, we questioned how chaotic electronic signals can be transformed into sound stimuli and explored their impact on brain activity using Electroencephalography (EEG). Our experiment involved 31 participants exposed to sounds generated from three processes from electronic implementations: signals from chaotic attractors, periodic limit cycles,and aleatory distributions. Our goal was to analyze characteristics and EEG signals to uncover the complex relationship between chaotic auditory stimuli and cognitive processes. Interestingly the chaotic stimuli caused a reduction in synchronization in the delta ((delta)) and theta ((theta)) frequency bands. We observed differences of up to 30 and 40%, primarily concentrated in the brain’s frontal areas. This desynchronization in (delta) and (theta) bands, seen in individuals, has implications for regulating irregular (theta) power in certain neural disorders. On the other hand, exposure to signals had mostly minimal effects on EEG readings. This research significantly contributes to our understanding of how the brain responds to stimuli derived from electronic systems. It sheds light on applications for modulating activity. Examining unpredictable sounds offers an understanding of the unique impacts of chaotic auditory inputs on brain activity, opening possibilities for further investigations at the crossroads of chaos theory, acoustics, and neuroscience.

Schizophrenia (SZ) is a serious mental disorder that can mainly be distinguished by symptoms including delusions and hallucinations. This mental disorder makes difficult conditions for the person and her/his relatives. Electroencephalogram (EEG) signal is a sophisticated neuroimaging technique that helps neurologists to diagnose this mental disorder. Estimating and evaluating brain effective connectivity between electrode pairs is an appropriate way of diagnosing brain states in neuroscience studies. In this study, we construct a novel image from multi-channels of EEG based on the fusion of three effective connectivity, partial directed coherence (PDC), and direct directed transfer function (dDTF) and transfer entropy (TE) at three consecutive time windows. Then, this image was used as input of five well-known convolutional neural networks (CNNs) through transfer learning (TL) to learn patterns related to SZ patients to diagnose this disorder from normal participants from two public databases. Also, the majority voting method was used to improve these results based on ensemble results of the five CNNs, i.e., ResNet-50, Inception-v3, DenseNet-201, EfficientNetB0, and NasNet-Mobile. The highest average accuracy, specificity and sensitivity to diagnose SZ patients from healthy participants were obtained using EfficientNetB0 through the Leave-One-Subject-out (LOSO) Cross-Validation criterion equal to 96.67%, 96.23%, 96.82%, 95.15%, 94.42% and 96.28% for the first and second databases, respectively. Also, as we suggested, the ensemble approach of EfficientNetB0, ResNet-50 and NasNet-Mobile increased the accuracy by approximately 3%. Our results show the effectiveness of providing fused images from multichannel EEG signals to the ensemble of CNNs through TL to diagnose SZ than state-of-the-art studies.

At present, at least 30% of refractory epilepsy patients in the world cannot be effectively controlled and treated. The suddenness and unpredictability of seizures greatly affect the physical and mental health and even the life safety of patients, and the realization of early prediction of seizures and the adoption of interventions are of great significance to the improvement of patients’ quality of life. In this paper, we firstly introduce the design process of EEG-based seizure prediction methods, introduce several databases commonly used in the research, and summarize the commonly used methods in pre-processing, feature extraction, classification and identification, and post-processing. Then, based on scalp EEG and intracranial EEG respectively, we reviewed the current status of epileptic seizure prediction research from five commonly used feature analysis methods, and make a comprehensive evaluation of both. Finally, this paper describes the reasons why the current algorithms cannot be applied to the clinic, summarizes their limitations, and gives corresponding suggestions, aiming to provide improvement directions for subsequent research. In addition, deep learning algorithms have emerged in recent years, and this paper also compares the advantages and disadvantages of deep learning algorithms with traditional machine learning methods, in the hope of providing researchers with new technologies and new ideas and making significant breakthroughs in the field of epileptic seizure prediction.