30703
Analysis of Neural Dynamics during Social Interaction with a Virtual Avatar in Autism
Social interactional neuroscience aims to explore the neural mechanisms underlying our social interactions, in order to better understand human behaviour and to reveal processes that are responsible for social impairment, one of the primary characteristics of neurodevelopmental disorders such as Autism Spectrum Disorder (ASD).
Objectives:
The aim of this study is to provide a more precise understanding on impaired reciprocal social interactions in ASD with the perspective of developing electroencephalography (EEG) based diagnostic methods in current child and adolescent psychiatric practice.
Methods:
In this study, neural dynamics of 40 typical controls (CTR) and 40 participants with ASD (Table 1) were recorded with high-density EEG (HD-EEG) during resting state and a social interactive task, where subjects coordinated their finger movements with a virtual avatar designed to behave like a human (Human Dynamic Clamp [HDC]). Neurophysiological data was analysed in order to compare the neural dynamics between the two groups. We computed Power Spectral Density (PSD) for each participant and investigated anatomo-functional differences in the usual frequency bands (Hz): Delta (1-4), Theta (4-7), Alpha (Low: 7-10, High: 10-13), Beta (13-30) and Gamma (30-45) and 8 brain lobes (Left frontal, Right frontal, Left parietal, Right parietal, Left occipital, Right occipital, Left temporal and Right temporal). Motor behaviour performances and verbal reports about the avatar’s behaviour were also assessed to compute sensorimotor and socio- cognitive metrics.
Results:
Social interaction with a virtual avatar revealed in both CTR and ASD groups a decrease in the Alpha band over the parietal lobes and an increase in the Beta and Gamma bands over the occipital lobe. However, we observed: 1) a greater High-Alpha suppression over the frontal lobe in ASDs, potentially reflecting inhibitory effects, and 2) higher increase in the Beta and Gamma bands over the occipital lobe in CTRs, potentially reflecting attentional effects; both observations are consistent with our previous findings. Behavioural analyses found small differences between groups, for sensorimotor scores (‘Phase Coordination’, ‘Amplitude Coordination’, ‘Task Realisation’) but more specially for socio-cognitive metrics (‘Intentionality’, ‘Theory of Mind’).
Conclusions:
These results support the development of social interactive tasks for the exploration of the neurophysiology of autism and its link to social cognition. The Human Dynamic Clamp allows to narrow the intra-individual variance while revealing inter-group differences. Our findings demonstrate that it can be used in clinical context and serves as a new instrument for the evaluation of on-line and off-line modes of social cognition. The HDC can also contribute to the improvement of existing therapeutic interventions for ASD that exploit virtual reality and social robotics.