| Project Detail |
The primate amygdala is a neural hub that processes computations for learning and memory, specifically when learning involves emotional, motivational, and reinforcement-based signals. This requires it to remain highly adaptive for changes in valence of environments and stimuli. Failures of such computations can lead to maladaptive behaviors and even psychopathologies such as PTSD and Anxiety. However, the core principles of amygdala function continue to elude the field. Recent studies suggest valence is processed in dedicated pathways, and we do not fully understand the mechanisms governing adaptive processing of valence and its reversal. Our overarching goal here is to elucidate the factors that underlie the dynamics of valence representation in amygdala circuits, and actively reverse valence to examine the impact on behavior. We develop a new framework using brain-computer-interface (BCI) and a closed-loop approach that allows us to guide changes in neural activity in the primate amygdala and modulatory networks. We test the hypothesis that coding properties of single amygdala neurons are dynamic, and examine how adaptive flexible coding is enabled by population activity. We unveil the parameters that govern this flexibility- timescales, directionality, population size, and its effective dimensionality. We then test how reversing representation of valence alters the animal response to learned stimuli, and use it to examine and manipulate aversive-biases in models of anxiety/trauma: generalization and exploration. Using high-density neural recordings in the primate amygdala, ACC, SI, with closed-loop behavioral paradigms and computational approaches, we will unveil a more direct (rather than correlative) role for the amygdala in the process of valence-based learning, and find the constraints that limit network adaptivity. Our findings in the primate brain will accelerate the design of closed-loop interventions to alleviate human psychopathologies. |
