Effect Investigation of Mask on Facial Expression Recognition Using Cerebral Evoked Potentials

There is little physiological evidence on mechanisms of facial expression recognition, which involves visual processing, feature extraction, emotion-processing, and cognitive integration. Therefore, this study aims to delve into the cerebral mechanisms underlying facial expression recognition using evoked potentials and understand how face masks affect these processes. We recorded the behavioral psychological and electrophysiological responses to facial expression stimuli: behavioral measures (accuracy and response time) and evoked potentials at 19 sites, which were measured in 12 subjects, corresponding to the type of facial expression and mask presence. The latencies and amplitudes of the five components (P1, N1, P2, N2, and P3) were analyzed at each site. For the behavioral measures, repeated two-way analysis of variance (ANOVA) demonstrated that accuracy was significantly affected by the type of facial expression (p < 0.001) and the presence of a mask (p < 0.01), with an interaction effect (p < 0.001). Response time was significantly affected by the type of facial expression (p < 0.001) and interaction (p < 0.01), however, it was not affected by the presence of a mask (p > 0.05). For evoked potentials, there were no significant differences in latency and amplitude for each site and component by type of facial expression (p > 0.05, one-way ANOVA). Nevertheless, there were significant differences in latency and amplitude for many sites and components with and without masks (p < 0.05, paired t-test). In addition, repeated two-way ANOVA revealed an interaction (p < 0.05) between N1 latency at F3, T3, C3, and Pz. In total, all these results suggest that the presence of a mask affects cognitive processing, and the presence of a mask for the type of facial expression affects central resources, both from a behavioral psychological and electrophysiological point of view.