Assessing the Development of Visual Shape Perception in Autism: Influence of Local Information

Background: Studies investigating visual perception in autism have identified altered perception at early (local) and higher (global) levels of visual analysis. However, much less is known about whether alterations at each level are related, and if so, whether such relationships differ at different periods of development. To assess whether a local attribute differentially affects global shape perception in autism as a function of development, sensitivity to Radial Frequency Patterns (RFPs; Wilkinson et al 1998; Grinter et al 2010) was assessed. RFPs are closed-contour shapes whose local spatial characteristics can be manipulated by creating “bumps” or RFs along their quasi-circular contours. Depending on the number of bumps, both local and global processing strategies can be targeted. Moreover, RFPs are known to target intermediate (global) levels of analysis, which precedes object perception. Local RFP attributes can also be manipulated by changing the physical attributes (i.e. luminance vs texture) defining their global contour, a manipulation that targets early level visual processes.      

Objectives: To assess whether the types of local information, mediated by early level perception, differentially affect intermediate level (global) visual perception in autism at different periods of development.  

Methods: 40 autistic and 44 non-autistic participants, matched for full-scale IQ and age, were placed into school-aged (7-12 years), adolescent (13-17 years) and adult (18-27) age groups. All participants were asked to discriminate between perfect circles and RFPs, whose contours (a) contained either 3, 5, or 10 bumps or RFs, and (b) were either luminance- or texture-defined. The size (or amplitude) of the bumps was varied: the larger the amplitude, the easier it was to discriminate a RFP from a perfect circle. Participants were asked which of two successively presented stimuli contained the RFP (target); the other was a perfect circle. RFP discrimination thresholds were measured using a method of constant stimuli and a 2-ATFC procedure. All participants had to complete a total of 6 experimental conditions.   

Results: Separate 2 (group) X 3 (age group) X 3 (# of RFs) mixed factorial ANOVAs were conducted for luminance- and texture-defined RFP conditions. For luminance-defined RFPs, no group-differences were identified at any RFP condition for any of the developmental periods assessed. For texture-defined RFPs, group-differences were identified for adolescents and adult groups, with decreased performance across both global (3 and 5 RFs) and local (10 RFs) RFP conditions in autism.    

Conclusions: The differential effect of type of local attribute (luminance vs texture) on global shape discrimination supports the hypothesis that decreased global perception in autism, when present, may have early (local) visual origins. Specifically, manipulating the complexity (from luminance to texture) of local attributes affected global shape perception to a greater extent in autism. Since the perception of certain types of objects, such as faces, is believed to be preceded by global shape analysis, it can be argued, based on the present results that alterations at early levels of analysis can in part contribute to the atypical perception of objects in autism, and this is especially evident later in development.