Neural Substrates of Simple and Complex Emotion Recognition in Autism: An fMRI study

Background: While some individuals with autism can succeed with simple emotion recognition, they show deficits with more complex displays of emotion (Baron Cohen, 1997). Therefore, a functional imaging study of simple and complex emotions provides the means to examine a graded response in individuals with autism. Objectives: To investigate functional brain activity during the recognition of simple and complex emotions in high-functioning adult males with autism. Methods: 12 high functioning males with autism and 12 matched controls gave informed consent and participated in the study. Stimuli were presented in a 3T MRI scanner via an overhead visor, with responses made using a hand-held response pad. Functional BOLD imaging was done using an echo-planar imaging sequence with TR/TE=2700/35 msec. We generated a battery of simple and complex emotional stimuli and a block design was used to randomly present 5 simple and 5 complex blocks of emotion matching trials. Each block consisted of 7 matching trials and was followed by a fixation period of 18900 msec. Brain Voyager software was used to identify regions of activation during each task and then subjected to a second-level random effects to identify group effects (Woods, 1996). Results: Individuals with autism showed greater activation in posterior and medial temporo-occipital regions and less robust activation of the fusiform gyrus to emotion face stimuli than controls. Additionally, while individuals with autism activated some regions associated with emotion processing (insula) during both simple and complex emotion recognition, controls showed heightened activation of emotion related regions as emotion recognition demands increased. Lastly, individuals with autism appear to use associative memory to support increased processing demands during complex emotion recognition. Conclusions: As compared to controls, individuals with autism engage different brain regions for simple and complex emotion recognition, and are increasingly distinguishable as emotion recognition demands increase.