​​​Talk from Trang-Anh Estelle Nghiem  - Stanford University, computational neuroscience

Short abstract:

Cognitive function and dysfunction emerge from intricate neural interactions spanning scales from single cells to the whole brain. Neuroimaging has made strides in characterizing the large-scale brain networks associated with cognitive functions and their alterations such as hallucinations and delusions in patients with psychosis. Yet, the causal cellular-scale mechanisms orchestrating brain network dynamics and their dysregulation in disorders remain unclear. In the case of psychosis, imbalance between excitation (E) and inhibition (I) has been hypothesized to underlie symptoms, but which brain regions may be affected by imbalance or whether it can cause the brain dynamical, sensory, and cognitive biomarkers of psychosis is still unknown. This gap in knowledge has crucially hampered the development of treatment strategies for psychiatric disorders specifically targeting the root causes behind symptoms.
 
In this talk, I will discuss how this major challenge can be addressed by leveraging biophysically informed modeling alongside computational approaches to neuroimaging data in my past, current, and proposed future work. First, I will discuss my research using biophysical modeling to reveal how cellular-level E/I imbalance shapes macroscopic-scale neural dynamics and responses to stimuli. My findings show that localized E/I imbalance can explain altered dynamics and function of the default mode network, an ensemble of brain regions affected across disorders including schizophrenia where psychosis is prevalent. Then, I will explore the default mode network’s cognitive role in representing behaviorally relevant information. Using state-space models inferred from neural data, I provide evidence for dynamical and distributed encoding of spatial behavior across default mode network regions. Finally, I will outline my proposed research plans to elucidate the localization and causal role of E/I imbalance in giving rise to atypical brain dynamics and sensory processing as well as psychosis symptoms of schizophrenia. The biophysically informed digital twin models developed in this work will form a basis to refine hypotheses and recommend treatment strategies effective for each patient. I will propose to conduct this planned research at Neurospin by applying for an ERC starting grant.