Robert Schmidt (BrainLinks-Braintools, University of Freiburg)

Abstract:
The neuromodulator dopamine has been implicated in both movement and learning. Evidence for its role in movement mostly comes from Parkinson's disease in which death of dopamine cells causes problems in patients to initiate and execute movements. The role of dopamine in learning was proposed based on similarities between dopamine cell activity patterns and reinforcement learning teaching signals (reward prediction errors). I will present several recent studies in which we used reinforcement learning models to describe dopamine dynamics and behavior in experimental data. In these studies we measured dopamine dynamics with electrophysiological recordings of dopamine neurons or with fast-scan cyclic voltammetry in the nucleus accumbens. Furthermore, we used optogenetics to mimick natural dopamine signals and study the effect on behavior. In summary, we found that phasic dopamine contributes to adaptive decision-making as both a teaching signal and an online modulator of vigor. Finally, I will provide a brief outlook on how these results relate to our future plans for modeling the impact of dopamine pulses on sensorimotor processing in the striatum.