Research Areas

Decoding the Content of Reward Prediction Errors

One way in which predictions about rewards are learned and updated is through prediction errors, which signal the difference between expected and experienced rewards. Recent experiments demonstrate that prediction error signals arising in the dopaminergic midbrain encode discrepancies in reward features beyond mere value, and are directly related to updates of those features in downstream prefrontal cortical areas. Ongoing efforts in the lab are aimed at deciphering whether prediction error signals themselves contain information about mispredicted events embedded in ensemble neural activity. These findings have implications for how the specificity of newly learned associations is conferred on downstream targets, and motivate future studies aimed at characterizing how sensory and thalamic networks integrate with mesocortical systems to guide optimal decisions.

Representation Mediated Learning

Sensory events can elicit vivid mental imagery of the rewards they have come to predict. When aversive or appetitive learning occurs for these events, to what extent does the learning carry over to the internally evoked associative representations? Animal model systems suggest that such mediated learning is indicative of overactive imagery, possibly modeling how hallucinations or delusions arise in human populations. We investigate the psychological and neurological underpinnings of representation-mediated learning in humans, aiming to shed light on how the brain discriminates between real and imagined consequences.

Conditioned Olfactory Hallucinations

Hallucinations are often intrusive, disruptive events in which an individual perceives a sensation when none is present. This effect is more robust in those with clinical disorders, including schizophrenia. In terms of sensory modality, the most commonly experienced hallucinations are auditory and visual. However, other sensory modalities, including olfaction, represent a significant portion of experienced hallucinations in both clinical and healthy populations. We aim to better understand experimentally-induced conditioned olfactory hallucinations in healthy adults. Findings could shed light on an underrepresented hallucination sensory modality experienced by a wide array of clinical populations and motivate future neuroimaging studies aimed at characterizing the underlying mechanisms driving false odor perception.

The Mediodorsal Thalamus and Odor-Guided Learning

A key function of the nervous system is to turn sensory input into meaningful behavioral output. In most sensory systems the thalamus serves as a projection point between sensory cells and their relevant sensory cortex. However in olfaction, sensory cells project to the olfactory bulb and then to the olfactory cortex. The olfactory network still has a thalamic input in the mediodorsal thalamus, though the region's exact role in olfaction remains elusive. We aim to address this knowledge-gap by combining an odor-guided learning task and ultra-high field fMRI techniques in healthy adults.

Thalamocortical Substrates of Stimulus Decoding and Behavior

Complex behaviors require the brain to associate abstract information with environmental stimuli in a context-dependent manner that goes beyond simple stimulus-motor responses. Evidence from rodent research has revealed that the thalamus is critical for promoting context-relevant representations in the frontal cortex while suppressing context-irrelevant ones, which suggests the thalamus plays a role in evaluating abstract information encoded within sensory input to help the frontal cortex deploy an optimal behavior. This project will use an olfactory-auditory attention task along with ultra high-field fMRI and transcranial magnetic stimulation to identify task-positive brain networks associated with this process and determine the role of the thalamus within them.