Characteristics of Facial Emotion Recognition in Children Aged 7-8 with High-Functioning ASD Using an Eye-Tracker

Objectives: The purpose of the present study was to investigate characteristics of peer’s facial emotion recognition (FER) in children with high-functioning autism spectrum disorder (HF-ASD) through gaze patterns; and to suggest implications through comparison with typically developing (TD children). Methods: The participants were 10 children with HF-ASD aged 7-8 years and 10 children with TD matched on chronological age. Children performed FER tasks consisting of two types of stimuli (static, dynamic) for four basic emotions (happy, sad, anger, surprise) and their eye movement were measured while they performed the task using an eye-tracker. The emotion recognition accuracy, proportion of fixation, and heatmap were analyzed, and correlation analysis was conducted. Results: The HF-ASD group performed significantly poorer than the TD group in overall emotion recognition and made significantly more errors only in anger. There was no statistically significant difference in emotion recognition between static and dynamic stimuli in both groups. There was no interaction effect between group and emotion type and areas of interest (AOI). Only the interaction effect between emotion type and AOI was significant in the TD group. The HF-ASD group tended to look more at the mouth than the eyes and look relatively more at non-feature area of the faces compared to the TD group. These gaze patterns seem to affect the accuracy depending on emotion types. Conclusion: Children with HF-ASD have deficits in FER and show gaze patterns different with TD peers. The results of this study suggest that a deficit in FER of children with HF-ASD is related to gaze pattern.