Objectives: This study investigated (1) whole-brain patterns of coherence during viewing of direct and averted eye gaze in ASD compared to neurotypicals (NT), and (2) the potential diagnostic utility of functional connectivity during gaze processing using discriminant analysis to differentiate groups.
Methods: 11 ASD and 8 NT adolescents and young adults passively viewed images of a character with direct or averted gaze while undergoing magnetoencephalography (MEG). A conditional button press ensured engagement. Synchronicity of the oscillating neurons was measured utilizing a Coherence imaging technique (ICA-MR-FOCUSS). Coherence values were calculated for each pair of 54 brain regions within each of 3 frequency bands: (1-15 Hz), beta (15-30 Hz), and low gamma (30-45 Hz). A between-groups t-test was conducted on each pair of brain regions for each frequency band within each condition (direct and averted). The False Discovery Rate method was controlled at 0.1 to correct for multiple comparisons. Of the pathways identified with significant between-group differences, coherence values with a difference > a standard deviation (SD) of 4 and without excessive within-group variance (≥ 0.15 SD) were entered into a linear discriminant analysis using leave-one-out cross-validation.
Results: Averted gaze in the 1-15 Hz band: ASD showed higher coherence between left parieto-occipital regions and right temporo-parieto-occipital regions and lower coherence between bilateral frontal and right fronto-temporo-parietal regions. In the beta band (associated with long range connectivity): ASD showed higher coherence between left parieto-occipital regions and bilateral temporo-occipital and left parietal regions. In the gamma band (associated with short range transmission): ASD showed higher coherence between bilateral temporo-parieto-occipital regions as well as bilateral parietal and orbitofrontal regions. In both beta and gamma bands: ASD showed lower coherence between bilateral frontal and right superior temporal regions. No significant differences were found during direct gaze. Of the pathways entered into the linear discriminant analysis, six were in the beta band and showed higher coherence in NT than ASD, and one was in the gamma band and showed higher coherence in ASD than NT. Linear discriminant analysis yielded 100% correct classification of both ASD and NT in the beta and gamma bands.
Conclusions: Results provide preliminary support for increased coherence in posterior, short range connectivity and decreased coherence in anterior to posterior, long- and short-range connectivity in ASD while viewing averted gaze. Neural synchrony between frontal cortex and pre- and postcentral gyri, and between frontal and superior temporal cortex in the beta band, and between angular and fusiform gyri in the gamma band were particularly discriminative. The high rate of correct classification of ASD vs. NT using coherence imaging in these pathways during passive viewing of averted gaze suggests it has potential diagnostic utility as a biomarker.
See more of: Brain Imaging
See more of: Brain Structure & Function