Variability of neural signaling is an important index of healthy brain functioning, as is signal complexity, which relates to information processing capacity. It is thought that alterations in variability and complexity may underlie certain brain dysfunctions. Here, resting-state fMRI was used to examine brain signal variability and complexity in male children and adolescents with and without autism spectrum disorder (ASD), a highly heterogeneous neurodevelopmental disorder. Variability was measured using the mean square successive difference (MSSD) of the time series, and complexity of these time series was assessed using sample entropy. A categorical approach was implemented to determine if the brain measures differed between diagnostic groups (ASD and typically developing (TD) groups). A dimensional approach was used to examine the continuum of relationships between each brain measure and behavioural severity, age, IQ, and the global efficiency (GE) of each participant's structural connectome, a metric that reflects the structural capacity for information processing. Using the categorical approach, no significant group differences were found for neither MSSD nor entropy. However, the dimensional approach revealed significant positive correlations between each brain measure, GE, and age. Further, negative correlations were observed between each brain measure and behavioural severity across all participants, whereby lower MSSD and entropy were associated with more severe ASD behaviours. These results reveal the nature of variability and complexity of fMRI signals in children and adolescents with and without ASD, and highlight the importance of taking a dimensional approach when analyzing brain function in ASD.
Modular networks of spiking neurons for applications in time-series information processing
