The neural processing of others' observed actions recruits a large network of brain regions (the action observation network, AON), in which frontal motor areas are thought to play a crucial role. Since the discovery of mirror neurons (MNs) in the ventral premotor cortex, it has been assumed that their activation was conditional upon the presentation of biological rather than nonbiological motion stimuli, supporting a form of direct visuomotor matching. Nonetheless, nonbiological observed movements have rarely been used as control stimuli to evaluate visual specificity, thereby leaving the issue of similarity among neural codes for executed actions and biological or nonbiological observed movements unresolved. Here, we addressed this issue by recording from two nodes of the AON that are attracting increasing interest, namely the ventro-rostral part of the dorsal premotor area F2 and the mesial pre-supplementary motor area F6 of macaques while they 1) executed a reaching-grasping task, 2) observed an experimenter performing the task, and 3) observed a nonbiological effector moving in the same context. Our findings revealed stronger neuronal responses to the observation of biological than nonbiological movement, but biological and nonbiological visual stimuli produced highly similar neural dynamics and relied on largely shared neural codes, which in turn remarkably differed from those associated with executed actions. These results indicate that, in highly familiar contexts, visuo-motor remapping processes in premotor areas hosting MNs are more complex and flexible than predicted by a direct visuomotor matching hypothesis.

中文翻译：

---

很大程度上共享用于生物和非生物观察到的运动的神经代码，但不适用于猴子前运动区的执行动作

其他人观察到的动作的神经处理会招募一个庞大的大脑区域网络（动作观察网络，AON），其中额叶运动区域被认为起着至关重要的作用。自从在腹侧前运动皮层中发现镜像神经元 (MN) 以来，人们一直认为它们的激活取决于生物而非非生物运动刺激的呈现，从而支持一种直接的视觉运动匹配形式。尽管如此，非生物观察到的运动很少被用作评估视觉特异性的控制刺激，从而使执行动作的神经代码与生物或非生物观察到的运动之间的相似性问题没有得到解决。在这里，我们通过从吸引越来越多的兴趣的 AON 的两个节点进行记录来解决这个问题，即猕猴背侧前运动区 F2 和内侧前补充运动区 F6 的腹侧部分，同时它们 1) 执行伸手抓握任务，2) 观察执行任务的实验者，以及 3) 观察非生物效应器在相同的上下文中移动。我们的研究结果表明，对观察生物运动的神经元反应比非生物运动更强，但生物和非生物视觉刺激产生高度相似的神经动力学，并依赖于很大程度上共享的神经代码，而这些神经代码又与与执行动作相关的神经代码显着不同。这些结果表明，在高度熟悉的情况下，与直接视觉运动匹配假设预测的相比，承载 MN 的前运动区域中的视觉运动重新映射过程更加复杂和灵活。