Objectives: To employ functional connectivity MRI (fcMRI) to investigate the organization of intrinsic functional networks subserved by pSTS in individuals with ASD and typically developing (TD) individuals.
Methods: Forty 9-18 year-old participants, 20 diagnosed with ASD and 20 age-, handedness-, and IQ-matched TD individuals, performed four runs of a visual search task in an event-related fMRI design (totaling 496 time points). Connectivity of networks existing independent of task effects can be detected in low-frequency fluctuations. Therefore, to isolate signals of interest, the timeseries were bandpass filtered (.008<f<.08 Hz), and the task paradigm, six motion parameters, and linear trends were included as regressors in a general linear model. To separate distinct signals within our regions of interest at the individual subject level, singular value decomposition of the fMRI timeseries from every voxel in pSTS was performed to obtain the singular timeseries that account for the greatest variance and thus are most representative of the functional network signals occurring within pSTS. Each of these timeseries was correlated with the whole brain and Fisher’s r-to-z' transformations were applied to the resultant maps. To separate the networks subserved by pSTS at the group level, principle component analysis was performed on z' connectivity maps. The functional connectivity of each network was further assessed within and between groups using t-tests.
Results: We found three networks subserved by pSTS in each hemisphere. The main network contained bilateral connections from pSTS with inferior frontal gyrus (IFG) and subcortical regions. The second pSTS network included connections with posterior cingulate cortex (PCC), precuneus, medial prefrontal cortex (mPFC), and IFG, possibly containing connectivity supporting processes related to theory-of-mind. The third network indicating connections between pSTS and area MT+, a region known for motion processing, was present in the TD group, but not in the ASD group. Direct-group comparisons of connectivity maps showed that in the ASD group several of the networks contained abnormally increased connectivity with superior frontal, middle frontal, and temporo-occipital regions and decreased connectivity in mPFC and superior parietal regions. Interestingly, in the ASD group, all three bilateral networks had largely overlapping connections, whereas much less overlap between networks was seen in the TD group.
Conclusions: We examined the functional connectivity of pSTS in each hemisphere and found that it participates in three distinct functional networks. The ASD group showed both abnormally increased and decreased connectivity. Generally, greater connectivity was reflected in correlations with regions outside of typical networks. Specifically, all three networks were less distinct in the ASD group and overlapped largely with one another, possibly due to reduced functional differentiation into separate, well-defined networks.