Cogn Neurodyn [JOURNAL]
Sun
200 papers
Unravelling emotions: exploring deep learning approaches for EEG-based emotion recognition with current challenges and future recommendations.
Abgeena A, Garg S
Cogn Neurodyn · 2025 Dec · PMID 41141240 · Full text
Emotion recognition (ER) is crucial for understanding human behaviours, social interactions, and psychological well-being. Electroencephalography (EEG) has emerged as a promising tool for capturing the neural correlates...
Emotion recognition (ER) is crucial for understanding human behaviours, social interactions, and psychological well-being. Electroencephalography (EEG) has emerged as a promising tool for capturing the neural correlates of emotions. This work is a systematic review of articles in ER using EEG signals. A total of 120 articles from 1041 articles were selected based on PRISMA guidelines using defined inclusion and exclusion criteria, published between 2018 and 2024. This article aims to provide an in-depth understanding of the current landscape of ER from EEG signals utilizing deep learning (DL). This review offers valuable guidance for researchers and practitioners seeking more refined and reliable emotion classification systems. To explore the effectiveness of DL models in EEG-based ER, several potential DL models, such as convolutional neural network, long short-term memory (LSTM), gated recurrent unit (GRU), hybrid bidirectional LSTM (BiLSTM), bidirectional GRU, and advanced DL models such as convolutional recurrent neural network and EEG-Conformer models are applied to two popular datasets, SEED and GAMEEMO, respectively, to depict the full process of ER. Additionally, the performance of DL models is also compared with the performance of basic machine learning (ML) models such as SVM, k-nearest neighbors, logistic regression, and boosting algorithms such as AdaBoost, XGBoost and LightGBM. Through extensive experiments and performance evaluations, the performance of different models when applied to the datasets mentioned above is compared. The accuracy, precision, recall, and F1-scores are analysed to determine the most effective model for EEG-based ER. The findings of this study demonstrate that the performance of hybrid DL models is more efficacious than that of ML models. The best-performing model (BiLSTM) classified the emotions, with an accuracy of 90.54% when applied to the GAMEEMO dataset. This research contributes to the growing body of literature on ER and provides insights into the feasibility of using EEG signals to understand emotional states, and presents a structured roadmap for future exploration. The findings can aid in the development of more accurate and reliable ER systems, which can have wide-ranging applications in psychology, social sciences, and human-computer interactions.
SH-StNN: prognostication of Alzheimer's disease based on search and hunt-based stacked deep convolutional neural network.
Mandawkar U, Diwan T
Cogn Neurodyn · 2025 Dec · PMID 41114389 · Full text
The conventional Machine Learning (ML) approaches for Alzheimer's disease (AD) detection using MRI images deployed the complex feature extraction strategies, consumed huge training time, and exhibited poor detection resu...
The conventional Machine Learning (ML) approaches for Alzheimer's disease (AD) detection using MRI images deployed the complex feature extraction strategies, consumed huge training time, and exhibited poor detection results. Particularly, Convolutional Neural Networks (CNNs) failed to capture long-range correlations from different brain regions, and suffer from overfitting issues. Hence, Select and Hunt Optimized Stacked Deep Convolutional Neural Network (SH-StNN) is proposed that automatically captures the intricate patterns associated with the brain structures, resulting in accurate detection for the effective AD detection. Architecturally, SH-StNN is constructed with the stacked-CNN layers, where RELU activation function is used. In this research, the Select and Hunt Optimization (SHO) algorithm is applied for medical image segmentation and effective classifier training, which optimizes the fifteenth layer of SH-StNN model. The experimental analysis demonstrates that the SH-StNN model shows improved accuracy of 98%, outperforming the existing techniques, such as Deep CNN by 13.17%, and CT-GAN by 10.81% for 80% of the training using the ADNI dataset. Additionally, the proposed SH-StNN model reports the accuracy of 96.73%, sensitivity of 96.90%, and specificity of 96.96% for the OASIS dataset.
Research on hippocampal positioning and navigation model based on energy fields.
Liu Y, Yan C, Tsuda I
Cogn Neurodyn · 2025 Dec · PMID 41104419 · Full text
Place cells in the hippocampus are crucial components of the brain's internal spatial positioning system, involved in constructing cognitive maps of the external environment for animals. However, many existing neuron mod...
Place cells in the hippocampus are crucial components of the brain's internal spatial positioning system, involved in constructing cognitive maps of the external environment for animals. However, many existing neuron models that simulate neural activities in the brain require extensive and complex computations. This study presents a place cell neural network model based on the Wang-Zhang model, using a neural energy coding approach. It quantitatively describes the attenuation pattern of place cell cluster firing power and constructs an energy field model. The model employs energy field gradients to address positioning and navigation tasks. Comparative experiments with the Hodgkin-Huxley (HH) model evaluate the navigation efficiency of rodents under different neuron models. The research shows that, compared to the HH model, the Wang-Zhang model has lower computational complexity and higher navigation efficiency. It rapidly constructs and updates cognitive maps, facilitating efficient pathfinding. Additionally, obstacle avoidance and detour experiments are performed using the Wang-Zhang model. Results demonstrate the model's ability for flexible navigation in dynamically changing mazes, validating the Wang-Zhang model and energy coding theory's unique functionality and robust advantages in neural modeling and information processing. This supports the effectiveness of energy coding in spatial memory and path exploration. Moreover, the additive property of neural energy provides significant advantages in neural modeling and computational analysis, offering a viable method for simulating large-scale neural networks and providing a theoretical basis for understanding the neurodynamic mechanisms of spatial memory.
Neural correlates of social influence in persuasion process: a hyperscanning EEG study on negotiation.
Ciminaghi F, Rovelli K, Acconito C … +1 more , Balconi M
Cogn Neurodyn · 2025 Dec · PMID 41098676 · Full text
Group decision-making requires integrating different perspectives through persuasion, which involves unidirectional social influence, and negotiation, which is a reciprocal interaction based on cooperation and competitio...
Group decision-making requires integrating different perspectives through persuasion, which involves unidirectional social influence, and negotiation, which is a reciprocal interaction based on cooperation and competition. While neuroscientific research has focused on identifying brain activations associated with these processes and their influencing factors, the impact of a prior persuasive dynamic on a subsequent negotiation task remains unexplored. This study examines whether engaging in a persuasive task, in which one individual has a role of social influence, affects neural activity during a subsequent negotiation. Using a hyperscanning paradigm with electroencephalography (EEG), frequency bands (delta, theta, alpha, beta and gamma) were analyzed in frontal, temporo-central and parieto-occipital regions in a sample of 26 participants. Results highlight distinct brain activation patterns between former persuaders and former receivers, with increased left-hemisphere delta activity and frontal theta and alpha activation in persuaders, while former receivers exhibited higher beta activity in the right parieto-occipital regions in the final stage of negotiation and higher gamma activity in right-lateralized regions. Overall, the study suggests that prior persuasive interactions shape subsequent negotiation at a neural level, influencing emotional, cognitive, and strategic engagement, with potential implications for understanding social dynamics in group interactions.
Correction: Differential impact of repetitive transcranial magnetic stimulation on alzheimer's disease symptomology: evidence from electrovestibulography.
Dastgheib ZA, Lithgow BJ, Moussavi ZK
Cogn Neurodyn · 2025 Dec · PMID 41078393 · Full text
[This corrects the article DOI: 10.1007/s11571-025-10310-5.].
[This corrects the article DOI: 10.1007/s11571-025-10310-5.].
Spatiotemporal transition of resting-state brain networks associates with human cognitive abilities.
Zhou L, Jiang Z, Chang Z … +2 more , Wang R, Wu Y
Cogn Neurodyn · 2025 Dec · PMID 41054555 · Full text
UNLABELLED: The brain is a dynamic system that continuously switches between different states. This brain state transition has significant functional consequences on human cognition, but its dynamic mechanism is rarely u...
UNLABELLED: The brain is a dynamic system that continuously switches between different states. This brain state transition has significant functional consequences on human cognition, but its dynamic mechanism is rarely understood. Here, we quantified the state transition by measuring the spatiotemporal reconfiguration of modular structure spanning time and space in the resting-brain functional networks. By integrating multimodal data, noise-driven large-scale dynamic model and meta-analysis, we found the significant relationship between state transition and brain evolution indicated by human accelerated regions (HARs) genes. This state transition was associated with diverse cognitive abilities, especially better executive control ability in the default mode network and control network. The resting-state brain showed a moderate degree of state transition at the whole-brain scale, but the regional heterogeneity of the transition was the highest, which functionally, was associated with the dynamic balance between segregation and integration, and structurally, was supported by hierarchical modules in brain structural connectivity. In addition, the high state transition among regions was supported by serotonin 1 A (5-HT) and dopamine (D) receptors. Our findings highlight the critical role of brain state transition in cognitive abilities and reveal the underlying dynamic mechanisms, offering new insights into the functional principles of the resting brain. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s11571-025-10347-6.
Graph theory methods for analyzing functional connectivity in multiple spike trains: application to data recorded from the visual cortex of a cat.
Masud MS, Nikolić D, Stuart L … +1 more , Borisyuk R
Cogn Neurodyn · 2025 Dec · PMID 41049270 · Full text
This study explores graph theory methods for analyzing the functional connectivity of multiple spike trains. We study simultaneously recorded multiple spike trains recorded from the visual cortex of a cat under different...
This study explores graph theory methods for analyzing the functional connectivity of multiple spike trains. We study simultaneously recorded multiple spike trains recorded from the visual cortex of a cat under different visual stimuli. To find the functional connectivity for a given visual stimulus we use the Cox method (Masud and Borisyuk, J Neurosci Methods 196:201-219, 2011). The application of graph theory methods for analysing the connectivity circuit, revealed that the functional connectivity of multiple spike trains is characterized by low density, long communication distances, and weak interconnectivity. Nevertheless, some spike trains also exhibit high degrees of centrality, including betweenness centrality, expansiveness coefficient, and attractiveness coefficient. Additionally, the analysis also identified significant motifs within the functional connections. Thus, our approach allows to describe the correspondence between the stimulus and functional connectivity diagram and compare functional connections under different stimuli.
Applications and interrelationships of brain function detection, brain-computer interfaces, and brain stimulation: a comprehensive review.
Xue Y, Chen Y, Wang F … +5 more , Zhao L, Li T, Gong A, Nan W, Fu Y
Cogn Neurodyn · 2025 Dec · PMID 41035905 · Full text
Brain-Computer Interface (BCI), Brain Function Detection (BFD), and Brain Stimulation (BS) are three pivotal technological domains in neuroscience and neuroengineering. Each plays a critical role in fundamental research,...
Brain-Computer Interface (BCI), Brain Function Detection (BFD), and Brain Stimulation (BS) are three pivotal technological domains in neuroscience and neuroengineering. Each plays a critical role in fundamental research, clinical applications, and human-computer interaction paradigms. Despite their distinct developmental pathways and application focuses, these technologies are frequently conflated or ambiguously referenced in both academic discourse and industrial practice, potentially leading to conceptual misinterpretations, suboptimal system designs, and clinical misapplications. Prior literature reviews have predominantly concentrated on BCI as a standalone subject, covering its historical evolution, specific neurophysiological signal modalities, or emergent technological trends. This manuscript's core contribution is critiquing the overuse of "passive BCI" (labeling feedback-absent monitoring as BCI). Through an application-oriented lens, it clarifies boundaries between BCI, BFD, and BS to resolve conceptual confusion. Further, the review interrogates the convergences and divergences among these modalities and critically evaluates the practical feasibility and challenges associated with their integrative deployment in clinical and experimental settings. Ultimately, this work aspires to provide a lucid, systematic, and conceptually coherent framework to support neuroscientific novices, interdisciplinary investigators, and clinical practitioners. By fostering precise comprehension and judicious utilization of BCI, BFD, and BS, it aims to propel their standardized advancement and enhance their translational impact across both research and clinical domains.
Self-repair mechanisms of spiking neuron-astrocyte networks in working memory under diverse injury conditions.
Mo B, Liu X, Li L … +4 more , Cheng S, Zhu Y, Yi M, Lu L
Cogn Neurodyn · 2025 Dec · PMID 41025128 · Full text
The injury to neurons and connection structures in the nervous system is a key factor leading to neurodegenerative diseases. Self-repair function refers to the innate capacity of the neuron-astrocyte network to partially...
The injury to neurons and connection structures in the nervous system is a key factor leading to neurodegenerative diseases. Self-repair function refers to the innate capacity of the neuron-astrocyte network to partially restore or maintain its function following injury, without external intervention. When the brain's nervous system is injured, how self-repair mechanisms work under various injury conditions and how to improve self-repair function remain unresolved. Through computational simulations of three distinct neurological injury scenarios, we investigated the self-repair function of spiking neuron-astrocyte networks in working memory tasks. Despite varying degrees of disruption of the network, all experiments (Self-Repair activated by synaptic connection injury, astrocytes injury, and internal noise interference) reveal that astrocytes can promote self-repair of the network during working memory tasks. Experiments on synaptic connection injury demonstrated that the network can maintain effective repair functionality under high injury conditions, which is associated with elevated calcium ion concentrations induced by increased glutamate release from presynaptic neurons. The modulation of astrocyte contributes to self-repair, and self-repair function decreases with increasing astrocyte injury. In addition, compared to the health network, internal noise interference has a small enhancement in the self-repair function of the network. Our findings elucidate the critical role of astrocyte-mediated signaling in maintaining network under different synaptic injury. This provides novel mechanistic insights into the threshold dynamics governing neuron network stability and early pathological transition in response to diverse neural injuries.
Complex nonlinear mechanisms for reduced response of hair bundle modulated by efferent nerve for protecting auditory function.
Cao B, Gu H, Wang R
Cogn Neurodyn · 2025 Dec · PMID 41025127 · Full text
UNLABELLED: As the ear's sensory receptors, hair cells detect sound vibrations via their hair bundles. A recent experimental report shows that efferent nerve activation can reduce hair bundle sensitivity, potentially pro...
UNLABELLED: As the ear's sensory receptors, hair cells detect sound vibrations via their hair bundles. A recent experimental report shows that efferent nerve activation can reduce hair bundle sensitivity, potentially protecting hair cells from loud sounds. In mammals, hair cells do not regenerate, making this protective mechanism crucial. However, the intrinsic dynamic mechanisms remain unknown. This paper integrates a model of the hair bundle, hair cell, and efferent nerve to reproduce the experimental observations initially. Then, the complex nonlinear mechanism for the reduced response is obtained with bifurcation analysis. The inhibitory synaptic current from the efferent nerve, when activated, causes a reduction in the membrane potential of the hair cell. This reduction further induces a decrease in the amplitude and an increase in the frequency of the mechanical oscillations of the hair bundle. Activation of the efferent nerve induces the appearance of weakened response or reduced sensitivity. Finally, feasible indicators to characterize sensitivity and measures to reduce sensitivity to enhance protection ability are obtained. The sensitivity is affected by oscillation patterns. Modulations of calcium concentration, conductance of electromechanical coupling current, or activation to efferent nerve are proposed to reduce sensitivity. The results present theoretical explanations to the protection function of the efferent nerve, potential measures to enhance the hearing protection ability, and a candidate coupling model to study these complex dynamics. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s11571-025-10349-4.
Construction and evaluation of an emotion-inducing video dataset towards Chinese elderly healthy controls and individuals with mild cognitive impairment.
Liang T, Yu J, Shi K … +9 more , Yao Y, Li J, Liu B, Wang W, Liu C, Qu L, Yin K, Xiang W, Li J
Cogn Neurodyn · 2025 Dec · PMID 41025126 · Full text
UNLABELLED: This work aimed to develop and validate an emotion-inducing video dataset for the Chinese elderly. The dataset was constructed by video collection, psychological evaluation, and elderly examination. 18 videos...
UNLABELLED: This work aimed to develop and validate an emotion-inducing video dataset for the Chinese elderly. The dataset was constructed by video collection, psychological evaluation, and elderly examination. 18 videos across six emotions (neutrality, sadness, anger, happiness, boredom, and tension) were selected for emotional induction. The effectiveness of the dataset was evaluated in 37 subjects, with two groups, 21 healthy controls (HC group) and 16 individuals with mild cognitive impairment (MCI group), who were assessed in a three-session experiment. Each session comprised one pretest and six emotion-inducing videos. The electrocardiogram (ECG) and electroencephalography (EEG) signals were synchronously recorded. After viewing each video, the subjects provided self-reports of discrete emotion labels, valence, and arousal scores using a modified Self-Assessment Manikin scale. Discrete emotion analysis, valence/arousal analysis, and ECG feature analysis were conducted by the ANOVA method. EEG feature analysis was assessed with a linear mixed-effects model. Discrete emotion analysis confirmed that happiness and sadness induced by the dataset show high agreement rates (e.g., happiness: HC 0.79, MCI 0.85 and sadness: HC 0.81, MCI 0.71), whereas boredom (HC 0.38, MCI 0.29) showed a comparatively lower consistency. Valence/arousal analysis revealed significant group differences for tension and boredom emotions. ECG feature analysis revealed significant differences in the baseline-normalized mean heart rate between HC and MCI groups in specific sessions. EEG feature analysis revealed that the MCI group exhibited higher relative band power values than did the HC group in the and bands. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s11571-025-10318-x.
Mtfsfn: a multi-view time-frequency-space fusion network for EEG-based emotion recognition.
Wang Z, Gao S
Cogn Neurodyn · 2025 Dec · PMID 41025125 · Full text
Over the recent years, emotion recognition based on electroencephalogram (EEG) has emerged as a prominent research area. Nevertheless, EEG signals present spatially discrete and non-stationary characteristics, to represe...
Over the recent years, emotion recognition based on electroencephalogram (EEG) has emerged as a prominent research area. Nevertheless, EEG signals present spatially discrete and non-stationary characteristics, to represent spatiotemporal information and extract more discriminative features from complex signals is still a challenge. This study proposed a multi-view time-frequency-space fusion network, referred to as MTFSFN. To effectively utilize complementary information from different frequency bands, we employ a frequency-domain attention mechanism to allocate weights to features of different frequency bands. A multi-view Transformer model was designed, integrating Transformer with two-dimensional positional embeddings to extract discrete spatial information. Following the fusion of multi-view features, we utilize LSTM to capture dynamic time-frequency-space relationships. Finally, a subject-independent leave-one-subject-out cross-validation strategy was used to validate extensively on three public datasets, DEAP, SEED, and SEED-IV. On the DEAP dataset, the average accuracies of valence and arousal are 78.64% and 77.42%, respectively. On the SEED dataset, the average accuracy is 86.91%. On the SEED-IV dataset, the average accuracy is 75.51%. The experimental results show that the proposed MTFSFN model achieves excellent recognition performance.
The impaired visual working memory of overweight and its intervention via six-week Tabata training: behavioral and event-related potential evidence.
Fu D, He Q, Wu T … +4 more , Wang X, Xiao M, Yuan J, Yan X
Cogn Neurodyn · 2025 Dec · PMID 41025124 · Full text
Overweight individuals often experience impairments in executive function, particularly working memory. Physical exercise has been shown to mitigate such cognitive decline and modulate brain activities. This study aimed...
Overweight individuals often experience impairments in executive function, particularly working memory. Physical exercise has been shown to mitigate such cognitive decline and modulate brain activities. This study aimed to investigate whether a six-week high-intensity interval (HIIT) Tabata exercise could improve working memory performance in overweight individuals and explore the associated neural mechanisms. To achieve this aim, two experiments were conducted. In Experiment 1, 20 overweight (Body Mass Index, BMI ≥ 24) and 20 health-weight university students completed the n-back task ( = 0 ~ 2) to assess working memory. Results confirmed that overweight participants exhibited lower accuracy (ACC) in the 2-back task compared with health-weight participants. Accordingly, in Experiment 2, another 40 overweight university students were randomly assigned into the training group (six-week HIIT Tabata) or control group (no physical exercise). All the participants performed the 2-back task with EEG recording at two points: before and after the six-week intervention (pre-test vs. post-test). Results showed that compared to pre-test, the training group showed higher accuracy at the post-test, whereas no such change was observed in the control group. Moreover, ERP results revealed a reduction in post-test P2 amplitude in the training group. Overall, this study demonstrates that being overweight negatively impacts working memory, while a six-week HIIT Tabata intervention may help alleviate these deficits, possibly through more efficient neural resource utilization.
R-CNN-TPOT: a new hybrid machine learning network for brain age prediction using EEG signal.
Almas S, Sosa PAV, Washakh RMA … +1 more , Waque RMU
Cogn Neurodyn · 2025 Dec · PMID 41025123 · Full text
Brain age refers to the significant changes in electroencephalogram (EEG) signals that occur as people age. The chronological age can be compared to the brain age to determine the variations from the normal ageing proces...
Brain age refers to the significant changes in electroencephalogram (EEG) signals that occur as people age. The chronological age can be compared to the brain age to determine the variations from the normal ageing process. With the rise of Machine Learning (ML), many brain age prediction methods have been developed using brain imaging. However, EEG-based approaches remain underexplored and have not utilized the Tree-based Pipeline Optimization Tool (TPOT). To subdue this problem, a novel hybrid ML technique is proposed for predicting brain age from EEG signals. The proposed method uses different features, such as spectral features, statistical features, frequency domain features and decomposition domain features. Additionally, a new ML approach called Regression-based Convolutional Neural Network-TPOT (R-CNN-TPOT) has been developed to perform the task of brain age prediction. Here, R-CNN-TPOT is obtained by combining the mathematical model of the Convolutional Neural Network (CNN) model and TPOT classification using regression modelling. In addition, the devised R-CNN-TPOT model provides better output with a Mean Absolute Error (MAE) of 0.033, Mean Square Error (MSE) of 0.063, R-squared of 15.456, and Root MSE (RMSE) of 0.251.
Innovative augmentation techniques and optimized ANN model for imagined speech decoding in EEG-based BCI.
Mohan A, Anand RS
Cogn Neurodyn · 2025 Dec · PMID 41025122 · Full text
Electroencephalogram (EEG) based Brain computer interface (BCI) emerges as a transformative technology with vast applications in neuroscience and rehabilitation. Imagined speech is the mental process of thinking and form...
Electroencephalogram (EEG) based Brain computer interface (BCI) emerges as a transformative technology with vast applications in neuroscience and rehabilitation. Imagined speech is the mental process of thinking and formulating words without vocalizing them through articulators. EEG signal is used to study imagined speech which can empower individuals with neurological impairments to communicate their thoughts effortlessly. The main challenge in decoding imagined speech is the nonstationary nature of EEG signals. Identifying robust features and scarcity of imagined speech datasets for properly training machine learning (ML) based algorithms is also a challenging task. The main objective of this study is to propose augmentation methods which mitigate data scarcity in EEG-based BCIs by introducing variations and strengthening model robustness through EEG data augmentation. The second objective is to propose a novel architecture capable of detecting variations in EEG signals for imagined speech datasets and show remarkable results. Seven diverse augmentation techniques are discussed, and the performance of the proposed model is analyzed in terms of accuracy, f1-score and kappa. The classification results are then compared with the case in which no data augmentation is used. The proposed model has shown remarkable accuracy of 91% for long words by using gaussian noise augmentation.
Dynamical insights from a heterogeneous whole-brain model of pre vs. post visual-motor task.
Wang T, Liu Y, Li B … +5 more , Xu R, Xu Y, Yang Y, Feng Y, Zhang L
Cogn Neurodyn · 2025 Dec · PMID 41020202 · Full text
UNLABELLED: Visual-motor task processing relies on neurovascular coupling (NVC), a neuro-hemodynamic interaction phenomenon. The brain's short-term effects following visual-motor tasks and the underlying mechanisms remai...
UNLABELLED: Visual-motor task processing relies on neurovascular coupling (NVC), a neuro-hemodynamic interaction phenomenon. The brain's short-term effects following visual-motor tasks and the underlying mechanisms remain largely unexplored. We developed a novel NVC-based dynamical model comprising multiple topologically coupled node units with intrinsic heterogeneity. Each node integrates a reverse neural mass model (RNMM) and a metabolic-hemodynamic model (MHM), interconnected via biophysically meaningful network connectivity matrix to enable cross-node interactions. The results show that, first, the model accurately replicated dynamic signatures during pre- and post-visual-motor task conditions, elucidating the NVC-mediated mechanism. Second, sustained elevation of transient metabolic-hemodynamic effects was observed in task-relevant regions (e.g., cuneus) post-task execution. Third, these short-term dynamical effects were jointly driven by NVC mechanisms and excitatory-inhibitory (E-I) balance regulation. In conclusion, our dynamical modeling approach elucidates the short-term effects jointly mediated by multiple mechanisms following visual-motor tasks, providing novel methodological and theoretical insights for understanding the cognitive mechanisms of brain function. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s11571-025-10351-w.
Effects of different frequencies of transcranial alternating current stimulation on emotion and emotion regulation.
Zhang S, Yu S, Li X
Cogn Neurodyn · 2025 Dec · PMID 41020201 · Full text
UNLABELLED: Emotion regulation is crucial in daily life, and transcranial alternating current stimulation (tACS) has the potential to improve it by modulating neural oscillations. In this study, 101 healthy adults were r...
UNLABELLED: Emotion regulation is crucial in daily life, and transcranial alternating current stimulation (tACS) has the potential to improve it by modulating neural oscillations. In this study, 101 healthy adults were randomized into four groups: frontal theta tACS, frontal alpha tACS, parieto-occipital alpha tACS, and sham control. Participants completed emotion regulation and facial Stroop tasks during stimulation, with emotional states assessed using the Positive and Negative Affect Scale (PANAS) and the Profile of Mood States-Short Form (POMS-SF) before and after stimulation. Physiological signals were also recorded during the stimulation. Repeated measures ANOVA analyzes were used for pre/post scale scores, task performance and physiological features. Results indicated that frontal theta tACS reduced negative emotions and improved reappraisal ability, whereas parieto-occipital alpha tACS showed comparable but non-significant effect. In contrast, frontal alpha tACS increased negative emotions and reaction times to disgust faces. These findings suggest that frontal theta tACS is a promising protocol for improving emotion regulation, and tACS may serve as a valuable tool for exploring neural mechanisms underlying emotional disorders. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s11571-025-10338-7.
Electroencephalography evidence of functional connectivity modulation and its correlation with bimanual visuomotor learning.
Phunruangsakao C, Hosoda C, Hayashibe M
Cogn Neurodyn · 2025 Dec · PMID 41020200 · Full text
Recent studies have shown that neuroplasticity related to sensorimotor adaptation can occur within short time frames, ranging from minutes to hours. However, it remains unclear whether bimanual training can induce simila...
Recent studies have shown that neuroplasticity related to sensorimotor adaptation can occur within short time frames, ranging from minutes to hours. However, it remains unclear whether bimanual training can induce similar effects. Therefore, the objective is to investigate immediate functional brain changes following brief bimanual visuomotor adaptation training. Node and edge-level electroencephalogram functional connectivity analysis and principal component regression were employed to examine changes related to visuomotor tracking task performance. The results revealed significant post-training improvements in bimanual performance, along with decreased node closeness centrality in the non-dominant right frontal and sensorimotor regions within the beta band, as well as in the right frontal, sensorimotor, and occipital regions within the gamma band. Edge-wise analysis indicated reduced beta- and gamma-band connectivity in the right hemisphere, aligning with the node-wise findings. Additionally, theta-band closeness centrality in the frontal, centroparietal, occipital, and temporal regions was positively correlated with bimanual performance, indicating a shift toward more centralized processing as performance increased. Principal component regression further demonstrated its predictive value for bimanual visuomotor performance. This study demonstrates that brief bimanual training elicits immediate functional connectivity changes associated with improved motor performance, particularly reduced right hemisphere beta/gamma connectivity and increased theta centrality. These findings highlight dynamic neural reorganization during bimanual adaptation. However, the interpretation of the results is limited by small sample size, EEG's low spatial resolution, and bias in functional connectivity estimation. These findings provide insights into adaptation mechanisms that could inform rehabilitation strategies for individuals with motor impairments.
Msst-eegnet: multi-scale spatio-temporal feature extraction using inception and temporal pyramid pooling for motor imagery classification.
Mishra R, Agrawal RK, Kirar JS
Cogn Neurodyn · 2025 Dec · PMID 40984876 · Full text
Motor imagery classification is an essential component of Brain-computer interface systems to interpret and recognize brain signals generated during the visualization of motor imagery tasks by a subject. The objective of...
Motor imagery classification is an essential component of Brain-computer interface systems to interpret and recognize brain signals generated during the visualization of motor imagery tasks by a subject. The objective of this work is to develop a novel DL model to extract discriminative features for better generalization performance to recognize motor imagery tasks. This paper presents a novel Multi-scale spatio-temporal network (MSST-EEGNet) to extract discriminative temporal, spectral, and spatial features for motor imagery task classification. The proposed MSST-EEGNet model includes three modules namely the inception module with dilated convolution, the temporal pyramid pooling module, and the classification module. Multi-scale temporal features along with spatial features are extracted using the inception block with the dilated convolution module. A set of multi-level fine-grained and coarse-grained features are extracted using a temporal pyramid pooling module. Further, categorical cross-entropy in combination with center loss is used as a loss function. Experiments are carried out on three benchmark datasets including the BCI Competition IV-2a dataset, the BCI Competition IV-2b dataset, and the OpenBMI dataset. The evaluation results shows that the proposed MSST-EEGNet model outperforms eight existing DL models in terms of classification accuracy for subject-specific and cross-session settings. It also outperforms eight existing DL models and six existing transfer-learning models for cross-subject setting. For the subject-specific classification the proposed MSST-EEGNet model achieved an accuracy of 0.8426 ± 0.1061, 0.7779 ± 0.0938, and 0.7365 ± 0.1477 on the BCI Competition IV-2a dataset, the BCI Competition IV-2b dataset, and the OpenBMI dataset respectively. For the cross-session setting, the proposed MSST-EEGNet model achieved an accuracy of 0.7709 ± 0.1098, 0.7524 ± 0.1017, and 0.6860 ± 0.0990 on the BCI Competition IV-2a dataset, the BCI Competition IV-2b dataset, and the OpenBMI dataset respectively. For the cross-subject setting, the proposed MSST-EEGNet model achieved an accuracy of 0.7288 ± 0.0730, 0.8161 ± 0.963, and 0.7075 ± 0.0746 on the BCI Competition IV-2a dataset, the BCI Competition IV-2b dataset, and the OpenBMI dataset respectively. Furthermore, a non-parametric Friedman statistical test demonstrates statistically significant superior performance of the proposed MSST-EEGNet model over the existing models.
Towards a transparent and interpretable AI model for medical image classifications.
Wen B, Wu Y, Daqqaq T … +1 more , Chaddad A
Cogn Neurodyn · 2025 Dec · PMID 40984875 · Full text
The integration of artificial intelligence (AI) into medicine is remarkable, offering advanced diagnostic and therapeutic possibilities. However, the inherent opacity of complex AI models presents significant challenges...
The integration of artificial intelligence (AI) into medicine is remarkable, offering advanced diagnostic and therapeutic possibilities. However, the inherent opacity of complex AI models presents significant challenges to their clinical practicality. This paper focuses primarily on investigating the application of explainable artificial intelligence (XAI) methods, with the aim of making AI decisions transparent and interpretable. Our research focuses on implementing simulations using various medical datasets to elucidate the internal workings of the XAI model. These dataset-driven simulations demonstrate how XAI effectively interprets AI predictions, thus improving the decision-making process for healthcare professionals. In addition to a survey of the main XAI methods and simulations, ongoing challenges in the XAI field are discussed. The study highlights the need for the continuous development and exploration of XAI, particularly from the perspective of diverse medical datasets, to promote its adoption and effectiveness in the healthcare domain. Our code is available at https://github.com/AIPMLab/XAI_-review-2024.
About
- Frequency
- Sun
- Papers found
- 200
- RSS feed
- Subscribe