Objectives: To investigate the neural mechanisms of biological motion recognition for evidence of different processing strategies between ASD and TD populations. In particular, to examine the effective connectivity between regions implicated in understanding the actions of others.
Methods: Participants in the study included an ASD group of 10 adult males and an age and IQ matched TD group. Every participant first performed a psychophysical task that involved recognizing the direction of locomotion of a point-light walker in visual noise (McKay, et al, 2009 Vision Research). This produced stimuli levels that equated to 50%(chance) and 84% correct performance thresholds for each individual. Following this all participants were scanned in a 3T Tim Trio scanner using a rapid, event-related design with three runs, each run showing 45 repetitions of every stimulus level. The experimental design was optimized for Granger Causality Mapping (GCM) so that each stimulus type had to cover a large enough number of continuous volumes for the analysis to compare over. Functional T2 weighted images were acquired with a TR of 1000ms. We collected 18 slices for each of 272 volumes at a resolution of 3mm x 3mm x 4.5mm slice thickness and dimensions 70 x 70 per image. Brainvoyager QX 1.10 was used for processing all stages of the data.
Results: An analysis of the behavioural data revealed that there was no difference between the stimulus thresholds for the ASD and TD groups. For the brain imaging analysis, a random-effect GLM was carried out on the group data. We first applied a mask generated from the regions derived from a preliminary experiment to restrict the analysis to only those regions that responded to these stimuli. Contrasts of brain activity for the 84% vs 50% thresholds revealed regions in the Inferior Parietal Lobule, Precentral Sulcus, Inferior Temporal Gyrus and Middle Frontal Gyrus for the TD group and the Middle Occipital Gyrus, Middle Temporal Gyrus, Fusiform Gyrus and Middle Frontal Gyrus for the ASD group. Each of the regions found from these contrasts were entered into the GCM analysis as seed regions. Results of this GCM analysis for the 84% condition indicated that the primary difference between TD and ASD groups was that the TD group engages processing between the temporal and parietal cortices while the ASD group engages areas only within the temporal cortex, including the Fusiform Gyrus, Lingual Gyrus and Parahippocampal Gyrus.
Conclusions: The TD group recruits processing in the dorsal stream to recognize biological motion while the ASD group relies on more extensive ventral processing. This is consistent with the ASD group recognizing biological motion in a more static template matching strategy while the TD group can engage further mechanisms of motion and social processing in the parietal cortex.