No abstract

Objective. Constructing an efficient human emotion recognition model based on electroencephalogram (EEG) signals is of great significance for realizing emotional brain computer interaction and improving machine intelligence. Approach. In this paper, we present a spatial-temporal feature fused convolutional graph attention network (STFCGAT) model based on multi-channel EEG signals for human emotion recognition. First, we combined the single-channel differential entropy (DE) feature with the cross-channel functional connectivity (FC) feature to extract both the temporal variation and spatial topological information of EEG. After that, a novel convolutional graph attention network was used to fuse the DE and FC features and further extract higher-level graph structural information with sufficient expressive power for emotion recognition. Furthermore, we introduced a multi-headed attention mechanism in graph neural networks to improve the generalization ability of the model. Main results. We evaluated the emotion recognition performance of our proposed model on the public SEED and DEAP datasets, which achieved a classification accuracy of 99.11±0.83% and 94.83±3.41% in subject-dependent and subject-independent experiments on SEED dataset, and achieved an accuracy of 91.19±1.24% and 92.03±4.57% for discrimination of arousal and valence in subject-independent experiments on DEAP dataset. Notably, our model achieved state-of-the-art (SOTA) performance on cross-subject emotion recognition task for both datasets. In addition, we gained an insight into the proposed frame by both the ablation experiments and the analysis of spatial patterns of FC and DE features. Significance. All these results prove the effectiveness of the STFCGAT architecture for emotion recognition and also indicate that there are significant differences in the spatial-temporal characteristics of the brain under different emotional states.

A 2-year-old girl, diagnosed with traumatic brain injury and epilepsy following car trauma, was followed up for 3 years (a total of 15 recordings taken at 0, 2, 3, 4, 5, 6, 7, 9, 10, 11, 12, 14, 19, 26, and 35 months). There is still no clear guidance on the diagnosis, treatment, and prognosis of children with disorders of consciousness. At each appointment, recordings included the child’s height, weight, pediatric Glasgow Coma Scale (pGCS), Coma Recovery Scale-Revised (CRS-R), Gesell Developmental Schedule, computed tomography or magnetic resonance imaging, electroencephalogram, frequency of seizures, oral antiepileptic drugs, stimulation with subject’s own name (SON), and median nerve electrical stimulation (MNS). Growth and development were deemed appropriate for the age of the child. The pGCS and Gesell Developmental Schedule provided a comprehensive assessment of consciousness and mental development; the weighted Phase Lag Index (wPLI ) in the β-band (13–25 Hz) can distinguish unresponsive wakefulness syndrome from minimally conscious state and confirm that the SON and MNS were effective. The continuous increase of delta-band power indicates a poor prognosis. Interictal epileptiform discharges (IEDs) have a cumulative effect and seizures seriously affect the prognosis.

Objective. One of the most significant features of the human brain is that it can dynamically reconfigure itself to adapt to a changing environment. However, dynamic interaction characteristics of the brain networks in naturalistic scenes remain unclear. Approach. We used open-source functional magnetic resonance imaging (fMRI) data from 15 participants who underwent fMRI scans while watching an audio–visual movie ‘Forrest Gump’. The community detection algorithm based on inter-subject functional correlation was used to study the time-varying functional networks only induced by the movie stimuli. The whole brain reconfiguration patterns were quantified by the temporal co-occurrence matrix that describes the probability of two brain regions engage in the same community (or putative functional module) across time and the time-varying brain modularity. Four graph metrics of integration, recruitment, spatio-temporal diversity and within-community normalised centrality were further calculated to summarise the brain network dynamic roles and hub features in their spatio-temporal topology. Main results. Our results suggest that the networks that were involved in attention and audio–visual information processing, such as the visual network, auditory network, and dorsal attention network, were considered to play a role of ‘stable loners’. By contrast, ‘unstable loner’ networks such as the default mode network (DMN) and fronto-parietal network tended to interact more flexibly with the other networks. In addition, global brain network showed significant fluctuations in modularity. The ‘stable loner’ networks always maintained high functional connectivity (FC) strength while ‘unstable loner’ networks, especially the DMN, exhibited high intra- and inter-network FC only during a low modularity period. Finally, changes in brain modularity were significantly associated with variations in emotions induced by the movie. Significance. Our findings provide new insight for understanding the dynamic interaction characteristics of functional brain networks during naturalistic stimuli.

No abstract

No abstract

Objective. Many recent studies investigating the processing of continuous natural speech have employed electroencephalography (EEG) due to its high temporal resolution. However, most of these studies explored the response mechanism limited to the electrode space. In this study, we intend to explore the underlying neural processing in the source space, particularly the dynamic functional interactions among different regions during neural entrainment to speech. Approach. We collected 128-channel EEG data while 22 participants listened to story speech and time-reversed speech using a naturalistic paradigm. We compared three different strategies to determine the best method to estimate the neural tracking responses from the sensor space to the brain source space. After that, we used dynamic graph theory to investigate the source connectivity dynamics among regions that were involved in speech tracking. Main result. By comparing the correlations between the predicted neural response and the original common neural response under the two experimental conditions, we found that estimating the common neural response of participants in the electrode space followed by source localization of neural responses achieved the best performance. Analysis of the distribution of brain sources entrained to story speech envelopes showed that not only auditory regions but also frontoparietal cognitive regions were recruited, indicating a hierarchical processing mechanism of speech. Further analysis of inter-region interactions based on dynamic graph theory found that neural entrainment to speech operates across multiple brain regions along the hierarchical structure, among which the bilateral insula, temporal lobe, and inferior frontal gyrus are key brain regions that control information transmission. All of these information flows result in dynamic fluctuations in functional connection strength and network topology over time, reflecting both bottom-up and top-down processing while orchestrating computations toward understanding. Significance. Our findings have important implications for understanding the neural mechanisms of the brain during processing natural speech stimuli.