Background: Research suggests that individuals with autism spectrum disorder (ASD) have difficulties in emotion recognition (ER), which could lead to social difficulties. ER can be enhanced through targeted interventions, but generalization to everyday functioning poses a challenge. Using dynamic multimodal emotional expressions for training may increase similarities to everyday situations. This pilot study investigated the feasibility of internet-based multimodal emotion recognition training (iMERAT) for adolescents with ASD.

Method: Eight adolescents with ASD and nine typically developing (TD) adolescents took part in the iMERAT intervention, which included brief online training sessions conducted each weekday during a 3-week period. Training was performed on dynamic facial, vocal and multimodal emotional expressions, with outcome feedback provided after each response. A survey was conducted to explore participants' experiences of the training. ER was measured pre- and post-training using a multimodal ER test.

Results: Participants reported that the training was moderately difficult, instructions were relatively easy to understand, and the duration of training was appropriate. Content analysis of open-ended responses suggested further adaptations, such as providing more explanations of emotions and further tailoring content and language for adolescents. ER increased from pre- to post-intervention, with large effect sizes for both ASD and TD adolescents.

Conclusion: Results suggest that the iMERAT intervention is feasible for adolescents with ASD. Gains in ER ability were observed, but the small sample size and lack of a control group render these findings tentative. Further research is required to assess the effectiveness of the iMERAT and possible impact on broader social skills.