Several other studies have found the precuneus to be involved in the representation of self. Taken together, these types of findings indicate disruption of self-related processing in autism associated with the precuneus and frontal regions. Findings of mean differences between autism and control groups in brain anatomy or brain activity have led more recently to classification studies in which participants are automatically classified as autistic or control based on such measures. Based on the structural grey matter anatomy measures, it was possible to classify the group membership with 85% accuracy. With the voxel-based morphometry approach, the accuracy was 90%. One study performed autism membership classification based on resting state connectivity data, producing an accuracy of 79%, whereas another study obtained an accuracy of 96%. There is apparently something distinctive about the brain structure and brain activation in autism. However, neither of these approaches relates a brain property to a specific type of concept or thought that is altered in autism. The current study examines whether such classification is possible based on the neural representations of interpersonal social CUDC-907 interactions, which might be expected to be altered in autism. In effect, the study seeks specific neurocognitive disruptions directly related to thought alterations and not simply biological markers of the thought disorder. We asked whether it is possible to distinguish autism from control participants based on their neural activation patterns during their consideration of various social interactions, examining whether the self components of social representations are altered in autism. In addition to relating altered neural activation patterns to social concepts, the study attempted to determine what anatomical alterations in autism might be associated with the psychological alterations in the conception of self. One theory of autism relates the disorder’s behavioral and brain activation symptoms to altered frontal-posterior anatomical connectivity in the cortex, compromising the communication bandwidth between frontal and posterior areas. The white matter tract that provides such connectivity between some of the main frontal and posterior midline regions involved in the representation of self is the cingulum bundle, whose structural properties can be measured noninvasively using magnetic resonance-based imaging of the diffusion of water molecules. An alteration in the representation of self could be due to the quality of this white matter tract. An a priori hypothesis was that the degree of alteration in the representation of self in individuals with autism would be related to the quality of their cingulum bundle. To examine this relation, diffusion images of this tract were obtained, in addition to the fMRI activation evoked by thoughts of various social interactions. Another hypothesis was that the degree of alteration in the representation of self in individuals with autism would be related to behavioral measures of various social abilities, such as face processing and Theory of Mind. To test this hypothesis, appropriate neuropsychological measures were acquired for participants with autism.