Graduate School of Education, The University of Tokyo Tokyo, Japan ; Japan Society for the Promotion of Science Tokyo, Japan ; Advanced Telecommunications Research Institute International Seika-cho, Japan.
Front Psychol. 2012 Dec 14;3:540. doi: 10.3389/fpsyg.2012.00540. eCollection 2012.
Sensory prediction error, which is the difference between actual and predicted sensory consequences, is a driving force of motor learning. Thus, appropriate temporal associations between the actual sensory feedback signals and motor commands for predicting sensory consequences are crucial for the brain to calculate the sensory prediction error accurately. Indeed, it has been shown that artificially introduced delays in visual feedback degrade motor learning. However, our previous study has showed that degraded adaptation is alleviated by prior habituation to the delay. Here, we investigate how the motor learning system accomplishes this alleviation. After the subjects habituated reaching movements in either 0- or 200-ms delayed cursor, visual rotation of 10° was imposed to the cursor with varying delay (0, 100, 200, or 300 ms) with each delay imposed in at least 1 out of 5-6 trials. Then, the aftereffect in the next trial was quantified to evaluate the adaptation response. After habituation to the 0-ms delayed cursor, the adaptation response was maximal when the visual feedback of the perturbation was provided with 0-ms delay and gradually decreased as the delay increased. On the other hand, habituation to the 200-ms delayed cursor alleviated the degraded adaptation response to the visual perturbation imposed during the 200-ms and longer delay (300 ms). However, habituation did not affect the adaptation response to the visual perturbation imposed during delays (0- and 100-ms delay) shorter than the habituated delay (200 ms). These results may be explained by assuming that habituation to the delayed feedback not only shifts the position of the hand predicted by motor command toward the delayed cursor positions, but also increases the degree to which the brain uses a certain amount of sensory prediction error to correct a motor command.
感觉预测误差是实际感觉后果与预测感觉后果之间的差异，是运动学习的驱动力。因此，准确计算感觉预测误差，大脑需要在实际感觉反馈信号和预测感觉后果的运动指令之间建立适当的时间关联。事实上，已经证明人为引入视觉反馈延迟会降低运动学习效果。然而，我们之前的研究表明，通过预先适应延迟可以减轻适应能力的下降。在这里，我们研究运动学习系统如何实现这种缓解。在适应 0 或 200ms 延迟光标后，向光标施加 10°的视觉旋转，并且在每个延迟中至少施加 5-6 次试验中的 1 次，视觉旋转的延迟分别为 0、100、200 或 300ms。然后，在下一次试验中量化滞后效应，以评估适应响应。在适应 0ms 延迟光标后，当视觉反馈与 0ms 延迟时，适应响应最大，随着延迟增加逐渐减小。另一方面，适应 200ms 延迟光标可以减轻在 200ms 及更长时间延迟（300ms）下施加的视觉扰动引起的适应能力下降。但是，适应不会影响比适应延迟（200ms）短的延迟（0-100ms）下施加的视觉扰动引起的适应响应。这些结果可以通过假设来解释，即适应延迟反馈不仅会将运动指令预测的手的位置向延迟光标位置移动，而且还会增加大脑利用一定量的感觉预测误差来纠正运动指令的程度。
