Farshchiansadegh Ali, Ranganathan Rajiv, Casadio Maura, Mussa-Ivaldi Ferdinando A
Sensory Motor Performance Program, Rehabilitation Institute of Chicago, Chicago, Illinois; Department of Biomedical Engineering, Northwestern University, Chicago, Illinois; and
Sensory Motor Performance Program, Rehabilitation Institute of Chicago, Chicago, Illinois;
J Neurophysiol. 2015 Jan 15;113(2):426-33. doi: 10.1152/jn.00249.2014. Epub 2014 Oct 22.
The goal of this study was to examine the reorganization of hand movements during adaptation to delayed visual feedback in a novel and redundant environment. In most natural behaviors, the brain must learn to invert a many-to-one map from high-dimensional joint movements and muscle forces to a low-dimensional goal. This spatial "inverse map" is learned by associating motor commands to their low-dimensional consequences. How is this map affected by the presence of temporal delays? A delay presents the brain with a new set of kinematic data, and, because of redundancy, the brain may use these data to form a new inverse map. We consider two possible responses to a novel visuomotor delay. In one case, the brain updates the previously learned spatial map, building a new association between motor commands and visual feedback of their effects. In the alternative case, the brain preserves the original map and learns to compensate the delay by a temporal shift of the motor commands. To test these alternative possibilities, we developed a virtual reality game in which subjects controlled the two-dimensional coordinates of a cursor by continuous hand gestures. Two groups of subjects tracked a target along predictable paths by wearing an instrumented data glove that recorded finger motions. The 19-dimensional glove signals controlled a cursor on a 2-dimensional computer display. The experiment was performed on 2 consecutive days. On the 1st day, subjects practiced tracking movements without delay. On the 2nd day, the test group performed the same task with a delay of 300 ms between the glove signals and the cursor display, whereas the control group continued practicing the nondelayed trials. We found evidence that to compensate for the delay, the test group relied on the coordination patterns established during the baseline, e.g., their hand-to-cursor inverse map was robust to the delay perturbation, which was counteracted by an anticipation of the motor command.
本研究的目的是在一个新颖且冗余的环境中，考察适应延迟视觉反馈过程中手部运动的重组情况。在大多数自然行为中，大脑必须学会将从高维关节运动和肌肉力量到低维目标的多对一映射进行反转。这种空间“逆映射”是通过将运动指令与其低维结果相关联来学习的。时间延迟的存在会如何影响这种映射呢？延迟为大脑呈现了一组新的运动学数据，并且由于冗余性，大脑可能会利用这些数据形成一个新的逆映射。我们考虑对一种新颖的视觉运动延迟的两种可能反应。在一种情况下，大脑更新先前学习的空间映射，在运动指令与其效果的视觉反馈之间建立新的关联。在另一种情况下，大脑保留原始映射，并学会通过运动指令的时间偏移来补偿延迟。为了测试这些不同的可能性，我们开发了一款虚拟现实游戏，在其中受试者通过连续的手部动作控制光标在二维空间中的坐标。两组受试者佩戴记录手指运动的仪器化数据手套，沿着可预测的路径追踪目标。19维的手套信号控制二维计算机显示器上的光标。实验连续进行两天。第一天，受试者练习无延迟的追踪运动。第二天，测试组在手套信号与光标显示之间存在300毫秒延迟的情况下执行相同任务，而对照组继续进行无延迟的试验。我们发现有证据表明，为了补偿延迟，测试组依赖于基线期间建立的协调模式，例如，他们的手到光标的逆映射对延迟扰动具有鲁棒性，这通过运动指令的提前预期得到抵消。
