Influence of Dopaminergically Mediated Reward on Somatosensory Decision-Making

Abstract

Reward-related dopaminergic influences on learning and overt behaviour are well established, but any influence on sensory decision-making is largely unknown. We used functional magnetic resonance imaging (fMRI) while participants judged electric somatosensory stimuli on one hand or other, before being rewarded for correct performance at trial end via a visual signal, at one of four anticipated financial levels. Prior to the procedure, participants received either placebo (saline), a dopamine agonist (levodopa), or an antagonist (haloperidol). Principal findings: higher anticipated reward improved tactile decisions. Visually signalled reward reactivated primary somatosensory cortex for the judged hand, more strongly for higher reward. After receiving a higher reward on one trial, somatosensory activations and decisions were enhanced on the next trial. These behavioural and neural effects were all enhanced by levodopa and attenuated by haloperidol, indicating dopaminergic dependency. Dopaminergic reward-related influences extend even to early somatosensory cortex and sensory decision-making. The rewards one receives during decision-making has a profound impact on learning. Much recent interest has focused on the role of the neurotransmitter dopamine in the basal ganglia for influencing learning and behaviour. Here, we ask whether reward can influence low-level sensory processing, for instance in primary sensory cortex, and how dopamine mediates this process. We show in humans that dopamine level, as manipulated with a dopamine agonist and antagonist in a double-blind placebo-controlled design, is involved in reward modulation of primary somatosensory cortex. Higher anticipated reward improved tactile decisions, and receipt of visual reward signals reactivated primary somatosensory cortex for the judged hand as measured using functional neuroimaging. After receiving a higher reward on one trial, somatosensory activations and decisions were enhanced on the next trial, suggesting that reward outcome provides a form of teaching signal that may be fed back to task-relevant sensory cortex. All these behavioural and neural effects of reward on somatosensory decision-making were strongly modulated by the availability of dopamine as the mediating neurotransmitter. These findings raise the tantalising new possibility that reward manipulations in conjunction with dopaminergic drugs might be used to enhance pathologically deficient or lapsed sensory processes, analogous to how rewards can be used to shape or correct behaviour.