Schedule of Events | Search Abstracts | Invited Symposia | Symposia | Poster Sessions | Data Blitz

High field strength fMRI reveals Reward Prediction Error-based learning signals in the Ventral Tegmental Area

Poster Session C - Sunday, March 8, 2026, 5:00 – 7:00 pm PDT, Fairview/Kitsilano Ballroom

Negar Yazdi¹ (), J Bruce Morton¹; ¹University of Western Ontario

Reward learning depends on reward prediction errors (RPEs) –the difference between expected and actual outcomes. In primates, dopaminergic neurons in the ventral tegmental area (VTA) and substantia nigra (SN) fire in proportion to RPE magnitude. Direct evidence of RPE-related activity in the human midbrain, however, remains limited due to the midbrain’s small size, deep location, and high susceptibility to physiological noise. Using high-resolution 7-Tesla fMRI, this study examined whether BOLD activity in dopaminergic midbrain nuclei varies with computationally defined RPEs. Ninety healthy adults (56 females; aged 18–30 years) completed an event-related reward-learning paradigm, in which participants learned stimulus–outcome associations. Data were preprocessed (motion correction special normalization), corrected for physiological noise (breathing, cardiac), and analyzed using SPM12 and PhysIO toolbox. At the first level, BOLD variability was modeled using a set of predictors scaled by the magnitude of the RPEs. Second-level analyses (FDR-corrected p < .05) revealed that activity in dopaminergic midbrain regions scaled as a function of positive prediction errors. Similar associations were observed in the activation of the striatum, hippocampus, salience, and frontoparietal networks. In dopaminergic nuclei and the hippocampus, these associations were pronounced early in learning but diminished as learning progressed. In the salience and frontoparietal networks, associations between BOLD amplitude and RPE remained consistent across the paradigm. These findings provide direct evidence of RPE-related BOLD activity in human midbrain dopaminergic nuclei and establish a foundation for investigating dopaminergic learning signals in humans using fMRI.

Topic Area: OTHER