Toward the detection of reduced emotion expression intensity: an autism sibling study

Expressing emotions through spontaneous facial expression is an important nonverbal social communication skill. In our study, we aimed to demonstrate that both children with autism spectrum disorder (ASD) and the non-ASD siblings of children with ASD have deficits in this skill.

In this study, we analyzed the six core facial emotion expressions of three distinct groups of children - those diagnosed with ASD (n = 60), non-ASD siblings (n = 60), and typically developed children (n = 60). To analyze facial expressions, we employed a computer vision program that uses machine learning algorithms to detect facial features and conducted an evidence-based task that involved assessing participants' ability to recognize facial emotion expressions.

Deficits in spontaneous emotion expression were shown in the children with ASD and in non-ASD siblings when compared with typically developed children. Interestingly, it was determined that these deficits were not related to the severity of the autism symptoms in the ASD group.

The results of the study suggest that computer-based automated analysis of facial expressions with contextual social scenes task holds potential for measuring limitations in the ability to express emotions, and they supplement the traditional clinical assessment of social phenotypical behavior deficits. This applies both to children with ASD and especially, to the non-ASD siblings of children with ASD. This study adds a novel approach to previous literature examining the emotion expression skills.