The Cognitive Neuroscience of Autism Spectrum Disorders

National Institute Of Mental Health

Investigators

Linked publications & trials

Abstract

This past year we have continued to evaluate the neural and behavioral characteristics of high-functioning individuals with an Autism Spectrum Disorder (ASD). We have been particularly interested in understanding aberrant eye movement patterns in ASD when confronted with dynamic social stimuli. W e have also used our work on using closed-loop neurofeedback in ASD as a model for demonstrating causal relationships between network functioning and behavior. Extensive study of typically developing (TD) and ASD participants have demonstrated that measures of eye tracking provide insights into social processing deficits in ASD, especially in conjunction with dynamic, naturalistic, freely-viewed stimuli. However, the question remains as to whether eye-gaze characteristics, such as preference for viewing specific facial features, can be considered a stable individual trait, particularly in those with ASD. Following up on our previous work (Ramot et al., Communications Biology, 2020), we assessed the stability and robustness of gaze preference for facial features as incremental amounts of movie data were introduced for analysis. To accomplish this we trained an artificial neural network to create an object-based segmentation of naturalistic movie clips as participants viewed multiple (25) Hollywood movie clips, each depicting a social interaction. As we evaluated combinations of one, three, five, eight, and 11 movie clips, gaze times on core facial features became increasingly stable at within-subject, within-group, and between-group levels. Using a number of movie clips deemed sufficient by our analysis, we found that individuals with ASD displayed significantly less face-centered gaze (centralized on the nose) but that viewing times did not significantly differ from typically developing participants when focusing on eyes or mouths. Our findings validate gaze preference for specific facial features as a stable individual trait and highlight the possibility of misinterpretation with insufficient data. Additionally, we propose the use of a machine learning approach to stimuli segmentation to quickly and flexibly prepare dynamic stimuli for analysis (Reimann et al., Autism Research, 2021). We recently applied the methods we developed for evaluating functional connectivity using fMRI data collected at rest in ASD to another development disorder; Downs syndrome (DS). Studies of resting-state functional connectivity in young people with DS have yielded conflicting results. Some studies have found increased neural connectivity while others have found a mix of increased and decreased connectivity. However, no studies have examined the relationship between whole-brain neural connectivity and network selectivity in DS (i.e., how distinct the networks are from each other). Our results revealed that whole-brain functional connectivity was significantly stronger in youth with DS (average age = 17.5 years; range 6-24, 13 of 19 subjects were female) compared to TD controls in widespread regions throughout the brain. Additionally, participants with DS had significantly reduced network selectivity compared to their TD peers. Exploratory behavioral analyses revealed that regions showing increased connectivity in DS predicted Verbal IQ, suggesting differences in connectivity may be related to verbal abilities. These results indicate that network organization is disrupted in youth with DS such that disparate networks are overly connected and less selective, suggesting a potential target for clinical interventions. Finally, following up on our previous work on closed-loop neurofeedback in ASD, we have been exploring the use of this training method for directly testing brain network causality. Specifically, we discuss how modulation of activity within a specific neural network using positive reinforcement may produce behavior change, thereby providing a direct test of network causality.