Principal Component Analysis of Eye-Tracking Data during Visual Perception of Human Faces in Adults and Children with Autism

摘要

Introduction: Autism is a pervasive developmental disorder, characterized by a triad of impairments: social communication problems, difficulties with reciprocal social interactions, and unusual patterns of repetitive behavior (Wing & Gould, 1979). This study was designed to explore particularities on visual perception of persons with autism. Thus, in order to investigate the gaze behavior of autistic individuals, statistical methods were performed. Methods and Materials: Principal Component analysis (PCA) was implemented to separate confused data sets of measured and estimated values of gaze behavior to make two distinct classes of observations ; a class of persons with autism and a second of healthy participants. Measured values were obtained from eye-tracking system which permitted to visualize sample neutral faces and recorded eye trajectory of each participant. A fractal study was also performed to find more discriminant parameters on which PCA was applied. Results and discussions: PCA showed that it is possible to classify in two groups human (adults and children) without and with autism.