Investigating the effects of perceptual processing on visual working memory in individuals with autism spectrum disorder : an event-related potential study
Chan, Yee Pei
Date of Issue2018-09-13
School of Social Sciences
Recent neuroimaging and electrophysiological research has shown that autism spectrum disorder (ASD) is a neural systems disorder characterised by atypical connectivity between brain networks (Minshew & Keller, 2010). Such atypical connectivity underlies an impairment in the visual working memory (VisWM) for complex information (Williams, Goldstein & Minshew, 2006). Although research has reported differences in perceptual processing as well as the working memory network of ASD (Williams et al., 2006), there is no published literature to our knowledge on the effects of such differences on the topography of effects of complex information processing in the VisWM of these individuals with ASD. While individuals with ASD have been shown to prefer local to gestalt processing (Minshew and Keller, 2010), there has been no published literature on perceptual closure processes. Perceptual closure, which involves the filling in of details to form a complete image, underlies gestalt processing (Kimchi, 1992). This study investigated how such differences in visual perceptual reasoning may impact the processing of complex information in the VisWM of individuals with ASD. Study 1 and Study 2 used an event-related potential (ERP) VisWM paradigm to examine differences in perceptual closure and the effect of coherence on object-to-scene binding in the VisWM of ASD. The effect of a perceptual training on perceptual closure for ASDs was then examined in Study 3. Nineteen ASD and 17 typically developing (TD) male adolescents from 13 to 16 years old participated in Study 1 and 2. Study 1 looked at the differences in perceptual processing between the adolescents with ASD and TD adolescents. Specifically, the perceptual closure processes involved in object recognition were examined. Perceptual closure involves the interplay between local and global processing, which has been found to be different in individuals with ASD. In this study, perceptual closure processes were indexed by closure negativity (NCL), an ERP component generated within the visual association cortex. Fragmented images were presented to the TD adolescents and adolescents with ASD. Adolescents with ASD were found to need a slightly higher level of completion before the fragmented object was recognised. When the ERP components were examined, adolescents with ASD and TD adolescents were found to have similar generation of early visual components (P1 and N1) and the NCL in the occipito-temporal regions. However, differences in prefrontal activations were found. While the TD adolescents showed prefrontal and posterior activations in the perceptual closure process, the individuals with ASD showed mainly temporal/parietal and posterior activations. Prefrontal regions have been shown to be involved in gist or low frequency processing. These areas have also been implicated in model selection and comparison processes. This is consistent with previous research that showed a reduced frontal-posterior connectivity in individuals with ASD. This engagement of prefrontal regions could explain the TD adolescents’ ability to recognise fragmented objects at a lower level of closure, where gestalt processing is critical. Having established that the interplay of global and local processing is present even in the earlier perceptual closure processes, Study 2 set out to investigate how such perceptual differences could impact on the ability of individuals with ASD to use coherence within an image to optimise their VisWM. In Study 2, slow-wave shifts showed that the VisWM network in adolescents with ASD and TD adolescents differed depending on the coherence of the scene. Although both groups showed better accuracy for intact than for scrambled scenes, TD adolescents recruited more higher-order attentional or cognitive processes when processing gestalt images. They also showed higher activation in the right brain regions, where gestalt processing occurs. Following the differences found in perceptual closure process and VisWM, Study 3 investigated if perceptual closure could be improved with the use of metacognitive strategies that encourage more gestalt processing and use of category comparisons. Seventeen male adolescents with ASD from 13 to 16 years old participated in Study 3. There were 9 participants in the training group and 8 participants in the control group. The differences in the two groups’ ability to identify fragmented images did not reach statistical significance. The two group did not differ in the number of correct images identified on their pre- and post-measures. However, the mean level of identification was significantly different before and after training for both groups. Both groups could identify the fragmented images at a less complete level. No significant differences in the two groups’ accuracy and reaction times were found on the VisWM tasks. The experiment would need to be replicated with a larger sample with an event-related potential design to investigate if differences in scalp activations are found following training. It could be that the differences may not be reflected at a behavioural level. Implications of the current findings suggest that interplay of global and local processing impacts on lower-level and higher-level visual processing. This in turn affects the VisWM processes and performance of individuals with ASD. The use of a perceptual training program for improving VisWM using coherence as a metacognitive strategy is an area that can be considered in ameliorating some of the difficulties resulting from the atypical perceptual processing of individuals with ASD.