EmoAnim: A Serious Game for Screening Children with Autism using Emotions in Animations

Abstract

Automatic autism screening is an important goal to avoid misdiagnosis of children with autism, especially in underdeveloped areas. That is why technology-based approaches for screening autism, especially game-based approaches, have been studied in the recent decade. This paper introduced a serious game called EmoAnim that utilizes animations to screen players' emotion recognition capabilities. The difference between EmoAnim and other emotion recognition games and tests was using animations rather than still images. Animations include both visual and auditory information that a child can use for emotion recognition. The game was tested on 25 children (10 children with autism and 15 normally developed children) aged 5 to 10. The results showed that typically developing children performed significantly better than children with autism in correct answer rate and response time. Moreover, children with autism spent more time in the game, specifically in the lower levels. Overall, the pattern of the gameplay between typically developing children and those with autism was different and identifiable by this game. Based on the study results, it can be concluded that this game can be used for screening purposes between children with autism and normally developing children. And the important feature of technology-based screening is that it might be used remotely without needing an expert presence. This would allow extending screening of autism in rural areas with inadequate healthcare support. This would facilitate screening autism, which would enable the family to start early rehabilitation interventions.