AI-based Computational Modeling Tools with Applications to Psychiatric Disorders
The ability to orient attention to dynamic changes in the environment is an essential cognitive skill. Aberrations in this ability are often observed in many psychiatric and neurological disorders, but the underlying neurophysiological mechanisms are not well understood. This project focuses on the development of novel biologically inspired artificial intelligence models for characterizing human brain function and dysfunction, with a special focus on improving our understanding of these mechanisms. We will develop novel recurrent neural network models of brain circuits underlying attention. We will then use these models to probe the nature of dysfunctional attentional circuits in patients with schizophrenia and autism. This will provide an in-silico platform to perform experiments and improve our understanding of the neurophysiological mechanisms underlying attention deficits. The findings will be applied to understand the empirical nature of aberrations observed in patients with schizophrenia and autism, compared to neurotypical controls. The studies will leverage the complementary expertise of the Stanford and French labs and help facilitate interactions between postdoctoral fellows, research scientists, and graduate students in the labs. Our collaborative studies will provide new tools for uncovering aberrant brain circuits underlying attention and cognitive control deficits in psychiatric disorders.