GCTNet: a graph convolutional transformer network for major depressive disorder detection based on EEG signals.

Objective.Identifying major depressive disorder (MDD) using objective physiological signals has become a pressing challenge.Approach.Hence, this paper proposes a graph convolutional transformer network (GCTNet) for accurate and reliable MDD detection using electroencephalogram (EEG) signals. The developed framework integrates a residual graph convolutional network block to capture spatial information and a Transformer block to extract global temporal dynamics. Additionally, we introduce the contrastive cross-entropy (CCE) loss that combines contrastive learning to enhance the stability and discriminability of the extracted features, thereby improving classification performance.Main results. The effectiveness of the GCTNet model and CCE loss was assessed using EEG data from 41 MDD patients and 44 normal controls, in addition to a publicly available dataset. Utilizing a subject-independent data partitioning method and 10-fold cross-validation, the proposed method demonstrated significant performance, achieving an average Area Under the Curve of 0.7693 and 0.9755 across both datasets, respectively. Comparative analyses demonstrated the superiority of the GCTNet framework with CCE loss over state-of-the-art algorithms in MDD detection tasks.Significance. The proposed method offers an objective and effective approach to MDD detection, providing valuable support for clinical-assisted diagnosis.

Categorical perception of facial expressions in individuals with non-clinical social anxiety.

BACKGROUND AND OBJECTIVES: According to the well-established categorical perception (CP) of facial expressions, we decode complicated expression signals into simplified categories to facilitate expression processing. Expression processing deficits have been widely described in social anxiety (SA), but it remains to be investigated whether CP of expressions are affected by SA. The present study examined whether individuals with SA had an interpretation bias when processing ambiguous expressions and whether the sensitivity of their CP was affected by their SA. METHODS: Sixty-four participants (high SA, 30; low SA, 34) were selected from 658 undergraduates using the Interaction Anxiousness Scale (IAS). With the CP paradigm, specifically with the analysis method of the logistic function model, we derived the categorical boundaries (reflecting interpretation bias) and slopes (reflecting sensitivity of CP) of both high- and low-SA groups while recognizing angry-fearful, happy-angry, and happy-fearful expression continua. RESULTS: Based on a comparison of the categorical boundaries and slopes between the high- and low-SA groups, the results showed that the categorical boundaries between the two groups were not different for any of the three continua, which means that the SA does not affect the interpretation bias for any of the three continua. The slopes for the high-SA group were flatter than those for the low-SA group for both the angry-fearful and happy-angry continua, indicating that the high-SA group is insensitive to the subtle changes that occur from angry to fearful faces and from happy to angry faces. LIMITATIONS: Since participants were selected from a sample of undergraduates based on their IAS scores, the results cannot be directly generalized to individuals with clinical SA disorder. CONCLUSIONS: The study indicates that SA does not affect interpretation biases in the processing of anger, fear, and happiness, but does modulate the sensitivity of individuals' CP when anger appears. High-SA individuals perceive angry expressions in a less categorical manner than the low-SA group, but no such difference was found in the perception of happy or fearful expressions.

Attentional resources modulate error processing-related brain electrical activity: Evidence from a dual-task design.

Attention plays an important role in the processing of error, but only a few studies have explored the relationship between them. The current study used a dual-task paradigm, combining the classic flanker task with a working memory load task, to explore how changes in the amount of attentional resources modulate error negativity (Ne) and error positivity (Pe). The results showed that the reduction of attentional resources overall caused a decrease in Pe amplitude, especially in the late stage of Pe, which had a significant diminution in amplitude. However, changes in the amount of attentional resources did not cause significant changes in the Ne amplitude. These results suggest that the early stage of error processing in the Ne time window is less affected by attention, but the Pe stage is regulated by attentional resources, especially in the late Pe stage.