Eye-Gaze Analysis of Facial Emotion Recognition and Expression in Adolescents with ASD

Abstract

Impaired emotion recognition and expression in individuals with autism spectrum disorder (ASD) may contribute to observed social impairment. The aim of this study was to examine the role of visual attention directed toward nonsocial aspects of a scene as a possible mechanism underlying recognition and expressive ability deficiency in ASD. One recognition and two expression tasks were administered. Recognition was assessed in force-choice paradigm, and expression was assessed during scripted and free-choice response (in response to emotional stimuli) tasks in youth with ASD (n = 20) and an age-matched sample of typically developing youth (n = 20). During stimulus presentation prior to response in each task, participants’ eye gaze was tracked. Youth with ASD were less accurate at identifying disgust and sadness in the recognition task. They fixated less to the eye region of stimuli showing surprise. A group difference was found during the free-choice response task, such that those with ASD expressed emotion less clearly but not during the scripted task. Results suggest altered eye gaze to the mouth region but not the eye region as a candidate mechanism for decreased ability to recognize or express emotion. Findings inform our understanding of the association between social attention and emotion recognition and expression deficits.

ACKNOWLEDGMENTS

We wish to acknowledge the support of the Virginia Tech’s College of Science, the Institute for Society, Culture, & Environment, and the Virginia Tech Center for Autism Research. We also greatly appreciate the help from Stephanie Roldan, who has aided in the development of the Matlab code to analyze the eye-tracking data.

FUNDING

This work was supported by grants from Virginia Tech Graduate Research Development Program and the Organization for Autism Research Graduate Research Grant.

Notes

¹ All results comparing TD and ASD participants remain unchanged when IQ was added as a covariate, aside from the interaction effect between group and emotion type on fixation to eye region for free-choice response task. This interaction effect is no longer significant when IQ is added as a covariate.

² These results are based on analyses per stimuli, and therefore there are incorrect responses for happy stimuli even though in Figure 3 there is almost perfect recognition of happy stimuli. Data for Figure 3 were analyzed per subject basis; data for Figure 6 were analyzed per stimuli basis.

Additional information

Funding

This work was supported by grants from Virginia Tech Graduate Research Development Program and the Organization for Autism Research Graduate Research Grant.