A Neural Population Model-based Characterization and Modulation of Neural Oscillations Underlying Cognitive Control in Healthy and Depressed/Anxious Human Subjects.

Cognitive control is a central element of human cognition that is impaired in several psychiatric diseases. It is closely associated with brain regions that are prominently dysregulated in disease such as the dorsal anterior cingulate cortex (dACC), the dorso-lateral prefrontal cortex (dlPFC), the orbitofrontal cortex (OFC), and the amygdala. Linking cognitive control to specific neural substrates and circuit-level interactions is critical for understanding how control disruptions lead to neurological and psychiatric diseases. This is work that is best done in humans, and so we will study cognitive control using intracranial local field potential recordings in human epilepsy patients. We will study cognitive control in patients who are having electrodes implanted for pre-surgical localization of epileptic seizures. These patients have electrodes in many of the brain regions involved in cognitive control, which is a unique opportunity to record from the human brain at a network level. Participants will perform laboratory tasks that measure cognitive control while we record the brain's electrical activity. The goal of this research proposal is to characterize and model cortical oscillations underlying the conflict resolution aspect of cognitive control in the healthy and anxious/depressed mental states to inform design of neuromodulation interventions for restoration towards healthy mental states. We will take three complimentary approaches to achieve this goal. 1) Using intracranial EEG recorded in epilepsy patients undergoing invasive monitoring for surgical evaluation, we will determine cortical neural oscillations underlying cognitive control in the framework of a conflict task. We will also use this to differentiate between normal and depressed or anxious subjects. 2) We will use a neural population cortical model to simulate and characterize feasible neural mechanisms underlying dlPFC and temporal cortical oscillations observed during conflict processing. We will use a data driven approach for estimating relevant model parameters. 3) We will use the model calibrated to individual subjects to design a cortical electrical stimulation paradigm to modulate the simulated oscillations towards a desired healthy state. We will then validate our model by testing model predictions in 4 human subjects and determine if we can successfully change both neural signals as well as associated behavior.

This is not a randomized controlled trial. All participants will receive brain recording, and all will receive stimulation once stimulation methods are set up. The randomization will be within subject: some blocks of the task will be performed under different conditions that emphasize different aspects of the control calculation. Some blocks will have brain stimulation, and some will not.