ST-SCGNN: A Spatio-Temporal Self-Constructing Graph Neural Network for Cross-Subject EEG-Based Emotion Recognition and Consciousness Detection

Pan, Jiahui; Liang, Rongming; He, Zhipeng; Li, Jingcong; Liang, Yan; Zhou, Xinjie; He, Yanbin; Li, Yuanqing

doi:10.1109/jbhi.2023.3335854

Cited by 11 publications

(3 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Meanwhile, this study adopted an offline learning strategy since the vision transformer-based models suffering from high resource occupancy. In addition, this study did not take cross-subject emotion recognition (He et al, 2021 ; Pan et al, 2023 ) into consideration, which may affect the applicability and universality of this study.…”

Section: Discussionmentioning

confidence: 99%

Music-evoked emotions classification using vision transformer in EEG signals

Wang,

Lian,

Cheng

et al. 2024

Front. Psychol.

View full text Add to dashboard Cite

IntroductionThe field of electroencephalogram (EEG)-based emotion identification has received significant attention and has been widely utilized in both human-computer interaction and therapeutic settings. The process of manually analyzing electroencephalogram signals is characterized by a significant investment of time and work. While machine learning methods have shown promising results in classifying emotions based on EEG data, the task of extracting distinct characteristics from these signals still poses a considerable difficulty.MethodsIn this study, we provide a unique deep learning model that incorporates an attention mechanism to effectively extract spatial and temporal information from emotion EEG recordings. The purpose of this model is to address the existing gap in the field. The implementation of emotion EEG classification involves the utilization of a global average pooling layer and a fully linked layer, which are employed to leverage the discernible characteristics. In order to assess the effectiveness of the suggested methodology, we initially gathered a dataset of EEG recordings related to music-induced emotions.ExperimentsSubsequently, we ran comparative tests between the state-of-the-art algorithms and the method given in this study, utilizing this proprietary dataset. Furthermore, a publicly accessible dataset was included in the subsequent comparative trials.DiscussionThe experimental findings provide evidence that the suggested methodology outperforms existing approaches in the categorization of emotion EEG signals, both in binary (positive and negative) and ternary (positive, negative, and neutral) scenarios.

show abstract

Section: Discussionmentioning

confidence: 99%

Music-evoked emotions classification using vision transformer in EEG signals

Wang,

Lian,

Cheng

et al. 2024

Front. Psychol.

View full text Add to dashboard Cite

show abstract

“…The results showed that the hybrid CNN-LSTM model achieved the highest accuracy of 94.17% on the raw DEAP dataset. Recently, graph neural networks (GNN) have shown excellent performance in EEG emotion recognition ( Zhang et al, 2022 ; Pan et al, 2024 ), which regard EEG signals as graph-structured data and extract high-level spatiotemporal information from EEG. Besides, some deep learning training strategies, such as domain adaptation ( He et al, 2022 ) and transfer learning ( Li J. et al, 2019 ), are highly favored especially in cross-subject EEG emotion recognition.…”

Section: Introductionmentioning

confidence: 99%

Emotion recognition based on microstate analysis from temporal and spatial patterns of electroencephalogram

Wei,

Li,

et al. 2024

Front. Neurosci.

View full text Add to dashboard Cite

IntroductionRecently, the microstate analysis method has been widely used to investigate the temporal and spatial dynamics of electroencephalogram (EEG) signals. However, most studies have focused on EEG at resting state, and few use microstate analysis to study emotional EEG. This paper aims to investigate the temporal and spatial patterns of EEG in emotional states, and the specific neurophysiological significance of microstates during the emotion cognitive process, and further explore the feasibility and effectiveness of applying the microstate analysis to emotion recognition.MethodsWe proposed a KLGEV-criterion-based microstate analysis method, which can automatically and adaptively identify the optimal number of microstates in emotional EEG. The extracted temporal and spatial microstate features then served as novel feature sets to improve the performance of EEG emotion recognition. We evaluated the proposed method on two publicly available emotional EEG datasets: the SJTU Emotion EEG Dataset (SEED) and the Database for Emotion Analysis using Physiological Signals (DEAP).ResultsFor the SEED dataset, 10 microstates were identified using the proposed method. These temporal and spatial features were fed into AutoGluon, an open-source automatic machine learning model, yielding an average three-class accuracy of 70.38% (±8.03%) in subject-dependent emotion recognition. For the DEAP dataset, the method identified 9 microstates. The average accuracy in the arousal dimension was 74.33% (±5.17%) and 75.49% (±5.70%) in the valence dimension, which were competitive performance compared to some previous machine-learning-based studies. Based on these results, we further discussed the neurophysiological relationship between specific microstates and emotions, which broaden our knowledge of the interpretability of EEG microstates. In particular, we found that arousal ratings were positively correlated with the activity of microstate C (anterior regions of default mode network) and negatively correlated with the activity of microstate D (dorsal attention network), while valence ratings were positively correlated with the activity of microstate B (visual network) and negatively correlated with the activity of microstate D (dorsal attention network).DiscussionIn summary, the findings in this paper indicate that the proposed KLGEV-criterion-based method can be employed to research emotional EEG signals effectively, and the microstate features are promising feature sets for EEG-based emotion recognition.

show abstract

“…and used algorithms such as Multilayer Perceptron Neural Networks and Random Forests to classify the level of consciousness ( Altintop et al, 2022 , 2023 ). Meanwhile, algorithms such as the synthetic minority oversampling technique (SMOTE) and the spatio-temporal self-constructing graph neural network (ST-SCGNN) also provide solutions to the data imbalance and cross-subject classification in consciousness detection research ( Chawla et al, 2002 ; Pan et al, 2023 ). Consciousness detection through machine learning is a possible approach.…”

Section: Introductionmentioning

confidence: 99%

A study on EEG differences between active counting and focused breathing tasks for more sensitive detection of consciousness

You,

Li,

et al. 2024

Front. Neurosci.

View full text Add to dashboard Cite

IntroductionIn studies on consciousness detection for patients with disorders of consciousness, difference comparison of EEG responses based on active and passive task modes is difficult to sensitively detect patients’ consciousness, while a single potential analysis of EEG responses cannot comprehensively and accurately determine patients’ consciousness status. Therefore, in this paper, we designed a new consciousness detection paradigm based on a multi-stage cognitive task that could induce a series of event-related potentials and ERD/ERS phenomena reflecting different consciousness contents. A simple and direct task of paying attention to breathing was designed, and a comprehensive evaluation of consciousness level was conducted using multi-feature joint analysis.MethodsWe recorded the EEG responses of 20 healthy subjects in three modes and reported the consciousness-related mean event-related potential amplitude, ERD/ERS phenomena, and the classification accuracy, sensitivity, and specificity of the EEG responses under different conditions.ResultsThe results showed that the EEG responses of the subjects under different conditions were significantly different in the time domain and time-frequency domain. Compared with the passive mode, the amplitudes of the event-related potentials in the breathing mode were further reduced, and the theta-ERS and alpha-ERD phenomena in the frontal region were further weakened. The breathing mode showed greater distinguishability from the active mode in machine learning-based classification.DiscussionBy analyzing multiple features of EEG responses in different modes and stimuli, it is expected to achieve more sensitive and accurate consciousness detection. This study can provide a new idea for the design of consciousness detection methods.

show abstract

ST-SCGNN: A Spatio-Temporal Self-Constructing Graph Neural Network for Cross-Subject EEG-Based Emotion Recognition and Consciousness Detection

Cited by 11 publications

References 47 publications

Music-evoked emotions classification using vision transformer in EEG signals

Music-evoked emotions classification using vision transformer in EEG signals

Emotion recognition based on microstate analysis from temporal and spatial patterns of electroencephalogram

A study on EEG differences between active counting and focused breathing tasks for more sensitive detection of consciousness

Contact Info

Product

Resources

About