Event Detail

Autism is a complex developmental disorder characterized by deficits across physical, social, and cognitive domains. Cognitive difficulties are found in executive function, “mind reading” abilities, the integration of information, attention, and the generalization of learned abilities to novel contexts. In addition, physical motor abnormalities, an increased prevalence of seizure disorders, motor stereotypies, and repetitive behaviors often accompany the diagnosis. The diversity of behavioral abnormalities exhibited in this disorder has prompted an almost equally diverse collection of psychological theories of autism, including Executive Dysfunction theories, Theory of Mind based accounts, and Weak Central Coherence theories.

This talk presents a general computational cognitive neuroscience model of interactions between the prefrontal cortex and the mesolimbic dopamine system which, when damaged, produces patterns of behavior that qualitatively and quantitatively match those observed in people with autism. The range of deficits captured by this approach is fairly broad, including aspects of executive dysfunction, stimulus overselectivity during conditioning, impaired implicit learning abilities, lexical disambiguation difficulties, and generalization problems in category learning. Thus, this computational account potentially offers a common neuroscientific explanation for diverse phenomena that have traditionally been explored within disjoint psychological frameworks.

Our model of interactions between the prefrontal cortex and the dopamine system will be described, and simulation results will be presented to demonstrate the ability of this model to capture the performance of both healthy and frontally damaged individuals on tasks involving working memory, cognitive control, and selective attention. Starting with this model of healthy performance, patterns of behavior matching those observed in people with autism will be produced by a simple disruption of dopamine modulation. Fits to behavorial data will be presented for a diverse collection of cognitive tasks, including Stroop, the Wisconsin Card Sorting Task, conditioning to multi-modal stimuli, a Serial Response Time Task, a lexical disambiguation task, and a prototype abstraction task.

David C. Noelle is Assistant Professor of Cognitive Science and Computer Science at the University of California, Merced. He received his Ph.D. in Cognitive Science and Computer Science from the University of California, San Diego. His research focuses primarily on computational cognitive neuroscience models of cognitive control, learning, and memory. The ongoing work reported in this talk is being conducted in collaboration with Trent Kriete, Ph.D., who is a postdoctoral fellow at the University of Colorado, Boulder.