Real-time measurement of neural activity with functional magnetic resonance imaging (rtfMRI) can be used as a source of feedback to guide volitional control (neuromodulation) of cognitive processing. Previous research has demonstrated broad success of rtfMRI-guided neuromodulation of cognitive processes. However, the neural mechanisms by which rtfMRI-guidance produces these outcomes remain largely unknown. Engineering control theory posits a small set of elements with specific component functions and causal associations necessary to achieve closed-loop control. The engine of this loop is the control law in which past control outcomes are evaluated to inform the selection of current control decisions. Cognitive control theory, meanwhile, posits numerous empirically justified measures by which the control law may evaluate outcomes as well as neural processing regions that may support these evaluation and selection functions, specifically, the medial and lateral prefrontal cortices. Dr. Keith Bush of the University of Arkansas for Medical Sciences will merge the predictions of these theories to model and explain brain activity observed during rtfMRI-guided neuromodulation. By advancing a mechanistic understanding of guided neuromodulation, this research will benefit multiple disciplines within cognitive science, psychology, and psychiatry that focus on self-control processes such as brain-computer interfaces, cognitive skills development, and cognitive therapeutics.
This research will conduct rtfMRI-guided neuromodulation experiments designed to facilitate volitional control of affective processing in healthy individuals. Affective processing will be volitionally modulated independently for the dimensions of valence and arousal to explore task-generality of the underlying closed-loop control system. In the first experiment, medial prefrontal cortex activity will be modeled with respect to multiple measures of control outcome (error, conflict, predicted error-likelihood, predicted response outcome, and expected value of control) to determine the most likely basis of its function. In the second experiment, control theoretic signals derived from this function will be used to locate brain regions implementing control signal selection and execution. Finally, the role of rtfMRI-guidance will be incorporated into the model to identify an anatomical locus of external feedback processing and a mechanistic prediction of how feedback informs the selection of control signals to improve closed-loop performance. The control system identified in this project would represent a biological basis of self-control processes across broad behavioral domains.
