Leveraging information theory to advance understanding of neurobiological mechanisms in autism spectrum disorder

The paper investigates the neurobiological mechanisms underlying Autism Spectrum Disorder (ASD) through the application of advanced data science techniques, including neuroimaging and computational modeling. Both global and regional abnormalities are considered, with particular attention to cerebellar contributions that may influence motor and cognitive symptoms associated with ASD. By employing machine learning classifiers and information-theoretic approaches, significant patterns in neural data are uncovered that contribute to a deeper understanding of ASD’s neurobiological underpinnings. Furthermore, integration of EEG connectivity studies with exploratory methods such as auditory neurofeedback and signal sonification—designed to modulate slow-wave (delta) activity—suggests preliminary potential for therapeutic application, though these approaches remain experimental. This comprehensive perspective aims to inform targeted interventions and enhance the understanding of ASD subtypes, ultimately contributing to improved outcomes for individuals affected by the disorder. • Explores neurobiological mechanisms of Autism Spectrum Disorder using advanced neuroimaging and computational modeling. • Applies information theory to neural systems to reveal atypical communication and connectivity patterns in Autism. • fMRI studies reveal atypical resting-state connectivity in Autism, especially in social and sensorimotor brain networks. • Computational models reveal diverse neural and social connectivity patterns in Autism using clustering and stochastic analyses.