Expectation modulates neural representations of valence throughout the human brain

• Published in: NeuroImage (Orlando, Fla.) Vol. 115; pp. 214 - 223
• Main Authors: Ramayya, Ashwin G., Pedisich, Isaac, Kahana, Michael J.
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 15.07.2015; Elsevier Limited
• Subjects: Adult; Algorithms; Anticipation, Psychological - physiology; Brain - anatomy & histology; Brain - physiology; Brain Mapping; Brain research; Decision making; Dopaminergic Neurons - physiology; Drug Resistant Epilepsy - surgery; ECoG; Electrodes; Electrodes, Implanted; Electroencephalography; Emotions - physiology; Feedback; Female; HFA; High frequency activity; Humans; iEEG; Intracranial electroencephalography; Learning - physiology; Male; Mesencephalon - physiology; Reinforcement (Psychology); Reinforcement learning; Reward; Valence; Value; Intracranial electroencephalography; High frequency activity; HFA; Valence; ECoG; Reinforcement learning; Value; Reward; iEEG
• ISSN: 1053-8119; 1095-9572; 1095-9572
• DOI: 10.1016/j.neuroimage.2015.04.037

The brain's sensitivity to unexpected gains or losses plays an important role in our ability to learn new behaviors (Rescorla and Wagner, 1972; Sutton and Barto, 1990). Recent work suggests that gains and losses are ubiquitously encoded throughout the human brain (Vickery et al., 2011), however, the extent to which reward expectation modulates these valence representations is not known. To address this question, we analyzed recordings from 4306 intracranially implanted electrodes in 39 neurosurgical patients as they performed a two-alternative probability learning task. Using high-frequency activity (HFA, 70–200Hz) as an indicator of local firing rates, we found that expectation modulated reward-related neural activity in widespread brain regions, including regions that receive sparse inputs from midbrain dopaminergic neurons. The strength of unexpected gain signals predicted subjects' abilities to encode stimulus–reward associations. Thus, neural signals that are functionally related to learning are widely distributed throughout the human brain. •We analyzed intracranial high frequency activity during reinforcement learning.•Expectation modulated reward-related neural activity in widespread brain regions.•These included regions that receive sparse inputs from midbrain dopamine neurons.•The strength of unexpected gain signals predicted subjects' learning during the task.•Neural signals related to learning are widely distributed throughout the human brain.