Assessing the Building Blocks of Spatial Perception in Autism Using a Tilt Discrimination Task

Background: As perceptual processing in autism is often described as detail-oriented, much can be learned about the spatial resolution of the visual system using stimuli and paradigms that directly assess the extraction of local orientation and spatial frequency. These early visual attributes, considered as the building blocks of spatial perception, are first encoded by orientation-specific mechanisms (spatial filters) operating within primary visual cortices. The tuning of these mechanisms - reflecting their spatial resolution - can be assessed by measuring the minimal amount of tilt that can be perceived relative to a reference orientation, across a variety of physical stimulus attributes (luminance- vs texture-defined) and spatial frequencies. This approach provides information about the extent to which “detailed” perception in autism stems from the increased tuning of mechanisms mediating early spatial perception.  

Objectives: To use a tilt discrimination task to assess whether the extraction of low-level spatial information is atypical among high-functioning individuals with autism.  

Methods: Fourteen participants with autism and 15 typically developing participants, matched for full-scale IQ and age (14-35 years), performed a tilt discrimination task. The participants were asked to indicate whether a grating was tilted to the left (counter-clockwise) or to the right (clockwise) relative to either a vertical or horizontal reference orientation. An adaptive staircase procedure was used to measure thresholds, defined by the minimal deviation (in degrees) needed to discriminate tilt orientation. Thresholds were obtained using gratings defined by luminance (with / without noise) and texture attributes of varying spatial frequencies (0.5, 1.0 and 2.0 cycles per degree (cpd)). The visibility of the gratings were equalized and kept constant throughout experimentation. Staircases ended after a 90 % confidence level, representing the threshold estimate falling within +/- 0.1 log units of the true measure, was obtained.

Results: A 2 (group) x 2 (reference orientation) x 3 (spatial frequency) mixed factorial ANOVA was conducted for each of the three attribute conditions (luminance with noise, luminance without noise, and texture) to evaluate whether groups differed with respect to tilt discrimination ability. For the luminance-defined, no-noise condition, the participants with autism had significantly lower (p < 0.05) tilt discrimination thresholds for horizontal gratings with a spatial frequency of 2 cpd. A significant group by orientation interaction (p < 0.05) revealed that the participants with autism were better at discriminating tilt relative to the horizontal axis for luminance-defined gratings with noise. No significant between-group differences and no significant interactions were found for texture-defined gratings.

Conclusions: The main finding of the present study was that participants with autism performed better than typically developing participants when discriminating tilt relative to a horizontal reference orientation for luminance-defined gratings with and without noise. This finding is evidence that the extraction of elementary, low-level spatial information is different among adolescents and young adults with autism, and consistent with increased tuning of early neural mechanisms mediating the extraction of elementary spatial information in autism. Such early alterations in autistic information processing may have effects on subsequent processing of higher-order stimuli, regardless of whether these stimuli are socially relevant.