Value coding of the striatal neurons

我们研究了尾状核神经元如何在工具性任务中编码时间折扣值。我们记录了尾状核中99个相位活动神经元。64个神经元表现出明显的线索相关活动。15个神经元（23%）与折扣结果值（DV）呈显著线性关系。在120个试验记录期间，超过15%的动机价值被折扣为奖励积累（即。例如，贬值）。然而，DV相关活动对这种奖励贬值并不敏感。这些结果表明，单个尾状核神经元的子集表示从预测的奖励大小和延迟持续时间计算的折扣结果值。尾状核折现值编码对贬值不敏感。这些价值编码似乎参与了基于价值的决策，而不是行为的动机调节。

We studied how caudate neurons code temporal discounted values in an instrumental task. We recorded the 99 phasically active neurons in the caudate nucleus. Sixty-four neurons showed significant cue-related activity. 15 neurons(23%) showed significant linear relation with discounted outcome value(DV). Within 120 trials recording periods, more than 15% of motivational value was discounted as reward accumulation(i. e., devaluation). However, the DV related activity was not sensitive to this reward devaluation. These results suggest that a subset of single caudate neuron represent the discounted outcome value computed from the predicted reward size and delay duration. The caudate discounted value coding is insensitive to devaluation. These value coding appears to be involved in value based decision making rather than motivational regulation of action.