Neural Bases of Hierarchical Shape Processing in Autism Spectrum Disorder

Background:  Enhanced visuospatial processing (Mottron et al., 2006) and Weak central coherence (WCC) (Frith & Happe, 2006) accounts argue that individuals with autism spectrum disorders (ASD) may focus on featural aspects of information more than the configural aspects. Nevertheless, there is increasing evidence that the “local bias” in ASD is not always present when other developmental factors (such as age and metrics of IQ) are taken into account (Pellicano et al., 2006). Furthermore, a number of experiments have found enhanced recruitment of frontal, parietal, and occipital brain regions in ASD, but studies reporting how these regions are interconnected during such visuospatial processing tasks in ASD is limited.

Objectives:  To obtain a greater understanding of brain circuitry involved in visuospatial functions in individuals with ASD.

Methods:  Fifteen children and adolescents with ASD and sixteen typically developing (TD) control participants (age 8-17 years) performed a hierarchical shape stimuli task (Navon, 1977) in the fMRI scanner. The task utilized a blocked design with trials that require the participant to name the larger shape composed of the smaller shapes (Global), and to name the smaller shapes that compose the large shape (Local). Data collected from a Siemens 3T Allegra scanner were analyzed using SPM12; functional connectivity analyses were performed using the CONN Toolbox, with 28 seed regions of interest (ROIs) selected from previous literature and from meta-analysis data. Connections were grouped and familywise error corrected using Network Based Statistics (NBS) in CONN. ANCOVAs were used to assess group differences in mean reaction time (RT) with age and full scale IQ (FSIQ) as covariates. 

Results:  While processing global shapes, ASD children displayed greater activity in the middle and superior temporal (MTG/STG), middle occipital, and lingual, superior and middle frontal gyri (SFG/MFG), and inferior parietal lobule (IPL), along with putamen and thalamus. A similar effect was found while identifying local shapes, with greater lingual, posterior cingulate, MTG, and SFG activation, along with putamen thalamus in ASD. ASD participants showed hyperconnectivity during local processing between the STG seed with the fusiform, inferior occipital, lingual, postcentral gyri and the thalamus. The thalamic seed revealed increased connections in ASD between the insula, inferior frontal, and postcentral gyrus. Faster RTs were found in the Local compared to the Global condition. Task accuracy did not differ by trial type or between ASD and TD individuals, but there was a significant main effect of FSIQ on task performance.

Conclusions:  Our functional results are consistent with some previous findings of increased activation among visual-spatial regions of the brain in ASD (Gadgil et al., 2013; Kana et al., 2013). Increased connectivity between these regions and the thalamus in ASD may suggest an altered pattern of perceiving and integrating visual stimuli in ASD. The behavioral results from this study do not support the hypothesis of enhanced perceptual processing for local stimuli in ASD, and suggest that performance for such tasks may be more influenced by other factors such as IQ. Future experiments with additional measurements are needed to explore this hypothesis further.