Department of Bioengineering, Temple University, Philadelphia, PA 19140, USA.
IEEE Trans Neural Syst Rehabil Eng. 2013 Mar;21(2):191-7. doi: 10.1109/TNSRE.2012.2237040. Epub 2013 Jan 9.
Sensorimotor coordination relies on the fine calibration and integration of visual, vestibular, and somatosensory input. Using virtual environments (VE) allows for the dissociation of visual and inertial inputs to manipulate human behavioral outputs. Our goal was to employ VE technology in a novel manner to investigate how head stabilization is affected by spatiotemporal properties of dynamic visual input when combined with passive motion on a linear sled. Healthy adults (n = 12) wore a head-mounted display during naso-occipital sinusoidal horizontal whole body translations while seated. Subjects were secured in a seat with a five-point harness, with the head free to move. Frequency and amplitude of sinusoidal input (i.e., inertial conditions) were set to create overlapping conditions of maximum acceleration (amax) or velocity (vmax). Four inertial conditions were combined with four visual conditions (VIS). VIS were created so that direction of optic flow either matched direction of passive motion or did not. The effect of near and far fixation distance within the VE was also tested. Head kinematics were collected with a three-axis gyro. Head stability showed a complex interaction dependent on changes in weighting of visual and inertial inputs that changed with the sled driving frequency. Inertial condition affected amplitude (p < 0.0000) and phase (p < 0.0000) of head pitch angular velocity. In the absence of visual input, head pitch velocity amplitude increased (p < 0.01). An interaction effect between inertial and VIS conditions on head yaw occurred in SW (p < 0.05). There was also a significant interaction of depth of field and inertial condition on amplitude (p < 0.001) and phase (p < 0.05) of head yaw velocity in SW, especially during high vmax conditions. We conclude visual flow can organize lateral cervical responses despite being discordant with inertial input. When using VE for rehabilitation, possible unintended, involuntary or reflexive motor responses that may not be present in traditional training environments should be taken into consideration.
感觉运动协调依赖于视觉、前庭和躯体感觉输入的精细校准和整合。使用虚拟现实（VE）可以分离视觉和惯性输入，从而改变人体的行为输出。我们的目标是采用一种新颖的 VE 技术来研究当与线性滑橇上的被动运动结合时，动态视觉输入的时空特性如何影响头部稳定性。健康成年人（n=12）在坐姿下佩戴头戴式显示器进行鼻枕窦正弦水平全身平移。受试者用五点式安全带固定在座椅上，头部可以自由移动。正弦输入的频率和幅度（即惯性条件）设置为产生最大加速度（amax）或速度（vmax）的重叠条件。四个惯性条件与四个视觉条件（VIS）相结合。VIS 的创建方式使得光流的方向要么与被动运动的方向匹配，要么不匹配。还测试了 VE 中近距和远距固定距离的影响。头部运动学由三轴陀螺仪收集。头部稳定性显示出复杂的相互作用，依赖于视觉和惯性输入权重的变化，而这些变化随滑橇驱动频率而变化。惯性条件会影响头部俯仰角速度的幅度（p<0.0000）和相位（p<0.0000）。在没有视觉输入的情况下，头部俯仰速度幅度增加（p<0.01）。在 SW 中，惯性和 VIS 条件之间存在头部偏航的相互作用效应（p<0.05）。景深和惯性条件对头部偏航速度幅度（p<0.001）和相位（p<0.05）的相互作用在 SW 中也很显著，尤其是在高 vmax 条件下。我们得出结论，尽管视觉流与惯性输入不一致，但它可以组织横向颈椎反应。在使用 VE 进行康复时，应该考虑到可能在传统训练环境中不存在的意外、无意识或反射性运动反应。
