The Forest and the Trees: Increased Sensitivity to Bilateral Symmetry In Autism

Background: The question of how local stimulus attributes are grouped into meaningful spatial patterns has not yet adequately been studied in autism. Bilateral symmetry, where half of a pattern is a mirror reflection of the other half, is a prototypical example of an ecologically-significant type of grouping. It exemplifies how local information is organized into highly salient and meaningful visual arrays involved in figure-ground segregation and object perception. Although symmetry perception is initiated by local comparison of individual elements on either side of a symmetrical axis, the extraction of its global symmetric configuration is contingent on integrative or global analysis, which in turn is found to solicit higher-order cortical visual areas in nonautistics. Typical individuals are often best at detecting symmetry oriented about a vertical axis, reflecting its ecological and social importance.

Objectives: To assess spatial grouping in autism by measuring sensitivity to bilateral visual symmetry oriented about vertical, oblique, and horizontal axes.

Methods: Fourteen autistic and 15 nonautistic participants matched for global IQ and age were asked to detect mirror symmetry oriented along randomly-presented vertical, oblique (45^o), and horizontal (90^o) axes. Patterns were composed of 500 dot-pairs (1000 total dots) and presented within a circular aperture for 250ms. Level of signal strength for symmetrical patterns was manipulated by varying the proportion of dots matched across the axis of symmetry (30%, 36.7%, 44.8%, 54.8%, 66.9%, 81.8% and 100%), such that 100% matching level meant the pattern was perfectly symmetrical. Symmetry detection thresholds were measured using a method of constant stimuli and a 2-alternative, temporal forced choice procedure by asking participants to detect which of two successively presented patterns contained symmetry (target), with one stimulus containing no symmetry signal (0% matching or non-target).

 Results: A 2 X 3 mixed-factorial ANOVA did not reveal an interaction between group and axis of symmetry orientation. However, a main effect of axis of orientation (p < 0.05) was identified, revealing that vertical symmetry (p < 0.05) was perceived most efficiently by both groups. In addition, a main effect of group (p < 0.05) was identified. Autistics demonstrated an advantage for symmetry perception, suggesting that autistics required less signal strength than did nonautistics, in order to detect visual symmetry across all axis orientations.  

Conclusions: A typically-manifested vertical advantage in the autism group suggests that mechanisms underlying symmetry perception in autism develop with, rather than in the absence of, innate and/or acquired sensitivity to environmental regularities that are important to nonautistics. However, autistics were atypically better at detecting visual symmetry across axes of orientation, which suggests that autistic perception is not characterised by a visuo-spatial integration deficit. Instead we suggest that under some circumstances, autistics can take advantage of an atypically parallel access to both local (symmetrical dot pairs) and global (oriented symmetrical patterns) spatial information. When genuine regularities exist within noisy visual arrays, autistics may sometimes see the forest and the trees.