Emotion classification on eye-tracking and electroencephalograph fused signals employing deep gradient neural networks

EEG Network Analysis of Depressive Emotion Interference Spatial Cognition Based on a Simulated Robotic Arm Docking Task

Depressive emotion (DE) refers to clinically relevant depressive symptoms without meeting the diagnostic criteria for depression. Studies have demonstrated that DE can cause spatial cognition impairment. However, the brain network mechanisms underlying DE interference spatial cognition remain unclear. This study aimed to reveal the differences in brain network connections between DE and healthy control (HC) groups during resting state and a spatial cognition task. The longer operation time of the DE group during spatial cognition task indicated DE interference spatial cognition. In the resting state stage, the DE group had weaker network connections in theta and alpha bands than the HC group had. Specifically, the electrodes in parietal regions were hubs of the differential networks, which are related to spatial attention. Moreover, in docking task stages, the left frontoparietal network connections in delta, beta, and gamma bands were stronger in the DE group than those of the HC group. The enhanced left frontoparietal connections in the DE group may be related to brain resource reorganization to compensate for spatial cognition decline and ensure the completion of spatial cognition tasks. Thus, these findings might provide new insights into the neural mechanisms of depressive emotion interference spatial cognition.

Multimodal EEG and Eye Tracking Feature Fusion Approaches for Attention Classification in Hybrid BCIs

Often, various modalities capture distinct aspects of particular mental states or activities. While machine learning algorithms can reliably predict numerous aspects of human cognition and behavior using a single modality, they can benefit from the combination of multiple modalities. This is why hybrid BCIs are gaining popularity. However, it is not always straightforward to combine features from a multimodal dataset. Along with the method for generating the features, one must decide when the modalities should be combined during the classification process. We compare unimodal EEG and eye tracking classification of internally and externally directed attention to multimodal approaches for early, middle, and late fusion in this study. On a binary dataset with a chance level of 0.5, late fusion of the data achieves the highest classification accuracy of 0.609–0.675 (95%-confidence interval). In general, the results indicate that for these modalities, middle or late fusion approaches are better suited than early fusion approaches. Additional validation of the observed trend will require the use of additional datasets, alternative feature generation mechanisms, decision rules, and neural network designs. We conclude with a set of premises that need to be considered when deciding on a multimodal attentional state classification approach.