Search

A key feature of learning in humans is flexibility in adjusting the weight of new information to update predictions. This flexibility can be computationally captured by changing the learning rate in a reinforcement‐learning model. Key components of reinforcement learning—such as prediction error (δ), learning rate (α), and reward feedback (r)—have been mapped to various brain areas. However, questions regarding the functional integration patterns in the human brain under the modulation of learning factors, and their interactions, remain unanswered. To investigate these phenomena, we first applied a reinforcement‐learning model with an adaptive learning rate and functional magnetic resonance imaging to simulate the individual's reward‐learning behavior. Psychophysiological interaction (PPI) analysis was then used to examine the functional interactions of the whole brain under the experimental condition of reward (r), the integration between reward and learning rate (α × r), and the integration between the prediction error and learning rate (α × δ) in a reward‐learning task. The behavior statistical analyses indicated that the model estimates of α and δ captured the participants' learning behavioral patterns of getting high reward, by changing α and δ for different difficulties and after getting different reward feedback. The PPI analysis results showed that motor‐related regions (including the supplement motor area, precentral gyrus, and thalamus) contributed to cognitive control processing regions (including the middle temporal gyrus, anterior and middle cingulate gyrus, and inferior frontal gyrus) by α × r. Finally, α × δ modulated the interaction between subregions of the striatum. [ABSTRACT FROM AUTHOR]

Abstract: Theories of the positive symptoms of schizophrenia hypothesize a role for aberrant reinforcement signaling driven by dysregulated dopamine transmission. Recently, we provided evidence of aberrant reward learning in symptomatic, but not asymptomatic patients with schizophrenia, using a novel paradigm, the Salience Attribution Test (SAT). The SAT is a probabilistic reward learning game that employs cues that vary across task-relevant and task-irrelevant dimensions; it provides behavioral indices of adaptive and aberrant reward learning. As an initial step prior to future clinical studies, here we used functional magnetic resonance imaging to examine the neural basis of adaptive and aberrant reward learning during the SAT in healthy volunteers. As expected, cues associated with high relative to low reward probabilities elicited robust hemodynamic responses in a network of structures previously implicated in motivational salience; the midbrain, in the vicinity of the ventral tegmental area, and regions targeted by its dopaminergic projections, i.e. medial dorsal thalamus, ventral striatum and prefrontal cortex (PFC). Responses in the medial dorsal thalamus and polar PFC were strongly correlated with the degree of adaptive reward learning across participants. Finally, and most importantly, differential dorsolateral PFC and middle temporal gyrus (MTG) responses to cues with identical reward probabilities were very strongly correlated with the degree of aberrant reward learning. Participants who showed greater aberrant learning exhibited greater dorsolateral PFC responses, and reduced MTG responses, to cues erroneously inferred to be less strongly associated with reward. The results are discussed in terms of their implications for different theories of associative learning. [Copyright &y& Elsevier]