Vasil'eva O N, Baginskas A
Institute of Higher Nervous Activity and Neurophysiology, Russian Academy of Sciences, Moscow, Russia.
Zh Vyssh Nerv Deiat Im I P Pavlova. 2003 Nov-Dec;53(6):681-96.
Amongst motor control and learning models, "A Cerebellar Model of Timing and Prediction" of A. Barto and J. Houk is the most interesting and physiologically well-grounded. Developing D. Marr's "The Theory of Cerebellar Cortex", this model proposed the important role in motor learning of the ability of Purkinje cells to change their activity level by the dendritic bistability mechanism. The aim of this investigation was to verify this idea in experiments with human learning of precise elbow flexion. The unsupervisual method of learning was used in order to guarantee the principal role of proprioception in training. The experiments were carried out in darkness to exclude the vision control. Subjects were asked to perform a precise horizontal elbow flexion as fast as possible and repeat this action from 30 to 50 times up to the point of complete movement acquisition (stable movement with the error in the range of 5% of a given flexion amplitude). The target point (a given angle of the horizontal elbow flexion) was not presented to the subjects in advance. Reaching the target point was indicated by a short light flash. During training, subjects learned to hit target point with the given precision. Kinematic characteristics of the movement (time change of elbow flexion angle, velocity, and acceleration) together with EMG of the flexor and extensor were recorded. The obtained results were in good agreement with J. Houk and A. Barto's hypothesis. Analysis of changes in the kinematic characteristics in the course of training revealed an asymmetric velocity profile and a fragmentary shape of acceleration profile at the beginning of learning. In the course of training, the acceleration profile transformed into biphasic curve with a single change in polarity. Thus, it acquired a characteristic shape of a plateau. Correspondingly, to the end of training, the character of the asymmetry of the velocity profile changed. No correlation was observed between the velocity parameters and movement precision. These features essentially distinguish the motor reactions under study from the common visuomotor coordinations. It is suggested that the amplitude and duration of the acceleration plateau reflect the intensity and time of inhibition of the descending activity of Purkinje cells as a result of bistability (in accordance with Houk and Barto's hypothesis).
在运动控制和学习模型中，A. 巴托和J. 霍克的《小脑计时与预测模型》是最有趣且在生理学上有充分依据的。该模型在发展D. 马尔的《小脑皮层理论》的基础上，提出浦肯野细胞通过树突双稳态机制改变其活动水平的能力在运动学习中具有重要作用。本研究的目的是在人类精确屈肘学习实验中验证这一观点。采用非视觉学习方法以确保本体感觉在训练中的主要作用。实验在黑暗中进行以排除视觉控制。要求受试者尽可能快地进行精确的水平屈肘动作，并重复此动作30至50次，直至完全掌握该动作（稳定动作，误差在给定屈肘幅度的5%范围内）。目标点（水平屈肘的给定角度）未提前告知受试者。到达目标点由短暂的闪光指示。在训练过程中，受试者学会以给定的精度击中目标点。记录了运动的运动学特征（屈肘角度、速度和加速度随时间的变化）以及屈肌和伸肌的肌电图。所得结果与J. 霍克和A. 巴托的假设高度一致。对训练过程中运动学特征变化的分析表明，在学习开始时，速度曲线不对称，加速度曲线呈碎片化形状。在训练过程中，加速度曲线转变为具有单个极性变化的双相曲线。因此，它呈现出平台的特征形状。相应地，到训练结束时，速度曲线不对称的特征发生了变化。未观察到速度参数与运动精度之间的相关性。这些特征将所研究的运动反应与常见的视觉运动协调本质地区分开来。有人认为，加速度平台的幅度和持续时间反映了由于双稳态导致的浦肯野细胞下行活动抑制的强度和时间（符合霍克和巴托的假设）。
