Entropy of Information in Clinical Assessment of Autism Spectrum Disorders

The absence of biomarkers poses a significant challenge to the early diagnosis of Autism Spectrum Disorders (ASDs) in children, particularly during routine newborn examinations. Consequently, post-birth years often rely on identifying difficulties in social interaction and other emerging factors to evaluate ASD in children. This study presents numerical findings examining social interaction and repetitive behaviors in both ASD and non-ASD cohorts. To illustrate these interactions, we employ two sets of random walkers, each representing distinct characteristics. One set, akin to children with autism, exhibits persistent and resistant dynamics, while the other, labeled as healthy, demonstrates diffusive behavior resembling the Elephant Random Walks (ERW) model. We analyze the influence of these two sets on each other's entropy variations using an information entropy framework. By examining paired interactions between walkers, we assess how these interactions impact the entropy fluctuations within each set. Surprisingly, our findings reveal that variations in the strength of interaction, represented by probability f, do not induce entropy changes in the group of random walkers simulating children with autism.