Analysis of Motor Signatures of Social Adaptation in Autism for Efficient Human-Centric Systems
This study addresses the identification of social adaptation differences in individuals with autism spectrum disorder (ASD) through movement behavior characteristics, aiming to inform precision medicine and human-computer interaction design. Using 3D motion capture data collected during a dance imitation task, the research compares movement patterns between autistic and neurotypical adults in both solitary and social contexts. It introduces the Social Context Sensitivity Index (SCSI), which—by quantifying how motor variability is modulated by social framing—proposes a novel potential motor biomarker for ASD. Movement consistency is assessed via dynamic time warping, and machine learning models are employed for group classification. Results reveal that neurotypical participants exhibit significantly increased upper- and lower-limb motor variability in social settings, whereas ASD participants maintain stable variability. The classifier achieves a balanced accuracy of 79.2%.