Cortical control of sequential motor behavior

我们许多熟练的行为不是单一的动作，而是一系列的动作。除了亚人类灵长类动物的神经记录外，人类的病变和成像研究表明，额叶皮层的内侧运动区对运动序列的产生尤为重要。不同区域的神经活动与序列表现的时空特征之间的确切关系尚不清楚。我们的一般论点是，更多的吻侧运动区域编码序列生产的全局方面，如序列订单或完整序列规范。而直接进入运动输出区域的神经元，如SMA、尾状扣带和运动皮层，则通过其定向特性的变化来反映序列。(1)确定运动皮层细胞的定向特性在多大程度上反映了过度学习运动序列的表现。假设是运动皮层中的细胞在序列产生过程中表现出定向特性的动态重组。(2)确定SMA和前SMA的神经活动与过度学习序列的时空关系。假设是，序列产生的全局方面将更常见地在SMA前编码，而SMA中的神经活动将反映神经元的定向特性和序列参数之间的相互作用。(3)记录过度学习序列执行过程中扣带运动区的神经活动。假设是，扣带运动区是一个完整的组成部分的皮层系统的顺序控制。使用的方法将是在训练有素的恒河猴的目标皮质区域进行标准的多电极记录，并使用适当的仪器进行行为控制和监测。我们的实验方法在几个重要方面不同于其他试图检查序列的神经相关性的尝试：我们将使用利用序列性能固有的空间维度的行为任务，我们特别强调记录神经元的空间特性的变化。

Many of our skilled behaviors involve not single actions but sequences of movements. Lesion and imaging studies in humans, in addition to neural recording in sub human primates have demonstrated that the medial motor areas in the frontal cortex are particularly important for the generation of motor sequences. What is less clear is the exact relation between neural activity in the different areas and the spatial and temporal features of sequence performance. Our general thesis is that the more rostral motor areas code for global aspects of sequence production such as serial order or complete sequence specification. Whereas neurons in areas with direct access to motor output such as the SMA, caudal cingulate, and motor cortex reflect sequences by changes in their directional properties. (1) To determine the extent to which the directional properties of cells in the motor cortex reflect aspects of the performance of over-learned motor sequences. The hypothesis is that cells in the motor cortex show a dynamic re-organization of directional properties during sequence production. (2) To determine the relation between neural activity in the SMA and pre-SMA and the spatial and temporal aspects of over-learned sequences. The hypothesis is that global aspects of sequence production will be more commonly coded in pre-SMA, while neural activity in SMA will reflect an interaction between the directional properties of the neurons and sequence parameters. (3) To record the neural activity in the cingulate motor areas during the performance of over-learned sequences. The hypothesis is that cingulate motor areas are an integral component of the cortical system for sequence control. The methods used will be those of standard multi-electrode recording in the target cortical areas of highly trained rhesus monkeys with appropriate instrumentation for behavioral control and monitoring. Our experimental approach differs from other attempts to examine the neural correlates of sequences in several important respects: we will use behavioral tasks which exploit the spatial dimension inherent in sequence performance and we place particular emphasis on changes in the spatial properties of the recorded neurons.