Psychophysiological study on how expectancy is acquired and modified

1. 我们研究了如何获得期望的神经元机制以及它是如何被修改的。我们用几种不同的奖励来训练猴子进行空间延迟反应任务和延迟反应时间任务。在延迟反应任务中，动物可以通过正确的任务表现获得奖励。同样的奖励持续进行了大约50次试验。当在不加热的情况下改变奖励时，动物在2到5次试验中获得了对新出现的奖励的期望。这种学习过程在动物的行为潜伏期和前额叶皮层的神经元活动中都可以观察到。在行为潜伏期和神经元活动之间，不可能确定哪个变化发生得更早。在延迟反应时间的任务中，动物被要求在一个go信号呈现后执行一个简单的按键反应。四个连续的试验组成一个块，并且在一个块内以相同的顺序给予4种不同的奖励：即，汗薯-葡萄干-卷心菜-无奖励（食物奖励）或橙汁-水-葡萄汁-无奖励（液体奖励）。因此，动物有可能知道在每次试验中会给予什么奖励，并且动物对每次试验的结果产生了期望。在灵长类动物眼窝额叶皮层中，存在三种与期望相关的神经元，即(1)监测动物面临的（1^<st>, 2^< and >, 3^<rd>或最后一次）试验的神经元，(2)参与判断当前试验将给予何种特定奖励的神经元，(3)与特定奖励预期相关的活动变化神经元。与奖励期望相关的眶额神经元的活动取决于奖励的动机意义，即动物越饥饿或口渴，奖励越受欢迎，其活动变化越大（或根据神经元的不同而不同）。

1. We investigated neuronal mechanisms of how expectancy is acquired and how it is modified. We trained monkeys on either spatial delayed response task or delayed reaction time task using several different kinds of reward.2. In the delayed response task, the animal could obtain the reward for the correct task performance. The same kind of reward was used continuously for about 50 trials. When the reward was changed without warming, the animal acquired the expectancy for the newly presented reward within 2 to 5 trials. This learning process was observed both in animal's behavioral latency and in neuronal activity of the prefrontal cortex. It was not possible to determine which changes occurred earlier between behavioral latency and neuronal activity.3. In the delayed reaction time task, the animal was required to perform a simple key press response after a go signal presentation. Four consecutive trials constituted a block, and 4 different kinds of reward were given always in the same order within a block : i.e., sweat potato - raisin - cabbage - no reward (food reward) or orange juice - water - grape juice - no reward (liquid reward). Thus, it was possible for the animal to know what reward would be given on each trial, and the animal developed the expectancy for the outcome in each trial. In the primate orbitofrontal cortex, there were three kinds of expectancy related neurons : i.e., (1) neurons that monitored which (1^<st>, 2^<nd>, 3^<rd>, or last) trial the animal was faced with, (2) neurons which were involved in judging what specific reward would be given in the current trial, (3) neurons which showed activity changes in relation to the expectancy of the specific reward. The activity of reward expectancy related orbitofrontal neurons depended on the motivational significance of the reward, that is, the more the animal was hungry or thirsty, and the more preferable the reward was, the larger (or the smaller depending on a neuron) the activity changes were.