Atypical Evidence Accumulation in Global Motion Decisions in Autism: Brain and Behavior

Friday, May 18, 2012: 11:00 AM
Grand Ballroom West (Sheraton Centre Toronto)
10:15 AM
C. E. Robertson1,2, C. Thomas2, D. Kravitz2, E. Dixon2, G. L. Wallace2, A. Martin2, S. Baron-Cohen3 and C. I. Baker2, (1)University of Cambridge, Cambridge, United Kingdom, (2)NIMH, National Institutes of Health, Bethesda, MD, (3)Autism Research Centre, University of Cambridge, Cambridge, United Kingdom

Numerous studies have reported a deficit in coherent motion perception in autism spectrum conditions (ASC). Research on motion perception has identified a neural circuit in which primary motion signals represented in middle temporal area (MT) are integrated in the banks of the lateral intraparietal sulcus (LIP) over time towards a decision-bound.  According to this model, a deficit in global motion perception should more strongly manifest with shorter than longer stimulus durations.  


To investigate neural and behavioral differences in rate at which motion signals are integrated towards a global percept in ASC.


36 adult participants (19 ASC) performed a forced-choice motion discrimination task manually indicating the global direction of motion (left/right) of a field of dots.  Stimulus duration varied between blocks (200/400/1500ms).  Coherence level (4-75%) and dot direction were randomly chosen on each trial. An additional 39 adult participants (20 ASC) also performed an event-related fMRI version of the motion coherence task.


Coherent motion perception thresholds were significantly higher in the ASC group (p<0.05) only at the shortest duration (200ms).  ASC and control performance at the longer durations were identical.  We replicate these results in the independent sample of individuals who participated in our fMRI study.  Turning to the functional data, we observed an overall reduction in the activation of the autistic MT across all coherence levels, but critically, this reduction was greater at the shorter stimulus durations. This reduction may lead to worse performance by slowing the formation of a decision variable and reducing its reliability.  Results of decoding analyses, in comparison with magnitude analyses, will also be discussed. 


We report a robust behavioral deficit in coherent motion perception in ASC when sensory integration time is limited, which is largely absent at longer stimulus durations.  These results point to atypical accumulation of motion signals in ASC: individuals with ASC require more evidence to reach a decision threshold than controls. Further, we have shown that this atypical accumulation is reflected in the reduced activation of the autistic MT. This result may provide insight into higher-order cognitive and social deficits that rely on visual integration, such as joint attention.

| More