Electrophysiological Correlates of Reinforcement Learning in the Human Ventral Tegmental Area
The ventral tegmental area (VTA) is the principal source of dopaminergic input to the human prefrontal cortex and is hypothesised to play a central role in reward-based learning. However, direct electrophysiological evidence in humans has been lacking. Here, using intracranial recordings obtained during an instrumental learning task in patients undergoing deep brain stimulation surgery for chronic cluster headache (n=13), we show that VTA local field potentials selectively encode rewarding outcomes compared with losses or neutral outcomes independently of whether behavioural learning occurred. In a subset of participants (n= 8) who learned to prefer a high-reward option, VTA responses to unexpected rewards were larger than to expected rewards. In two participants who sufficiently explored both options, VTA activity reflected the expected value of the chosen option during decision-making. These findings provide direct electrophysiological evidence that the human VTA encodes reward-related signals consistent with reinforcement learning mechanisms described in animal models.