Byagowi Ahmad, Moussavi Zahra
Department of Electrical and Computer Engineering, University of Manitoba, Winnipeg, MB R3T 5V6, Canada.
Annu Int Conf IEEE Eng Med Biol Soc. 2012;2012:4812-5. doi: 10.1109/EMBC.2012.6347070.
Virtual reality (VR) experiments are commonly used to assess human brain functions. We orient ourselves in an environment by computing precise self-to-object spatial relations (egocentric orientation) as well as object-to-object spatial relations (allocentric orientation). Egocentric orientation involves cues that depend on the position of the observer (i.e. left-right, front-behind), whereas allocentric orientation is maintained through the use of environmental features such as landmarks. As such, allocentric orientation involves short-term memory, whereas egocentric orientation does not. This paper presents a Virtual Reality Navigational (VRN) experiment specifically designed to assess egocentric spatial cognition. The design aimed to minimize the effect of spatial cues or landmarks for human navigation in a naturalistic VR environment. The VRN experiment designed for this study, called the Virtual House, is a symmetric three story cubic building, with 3 windows on each side on every floor, and one entrance on each side of the building. In each trial, a window is marked by a pseudo-random sequence as the objective. The marked window is shown to the participant from an outdoor view. The task is to reach the objective window using the shortest path through the building. The experiment entails 2 sets of 8 trials to cover all possibilities. The participants' performance error is measured by the difference between their traversed distance trajectory and the shortest natural distance (calculated using the VR engine), normalized by the shortest distance, in each trial. Fifty-two cognitively healthy adults participated in the study. The results show no learning effect during the 16 trails, implying that the experiment does not rely on short-term memory. Furthermore, the subjects' normalized performance error showed an almost linear increase with age, implying that egocentric spatial cognition ability declines with age.
虚拟现实（VR）实验常用于评估人类大脑功能。我们通过计算精确的自我与物体空间关系（自我中心定向）以及物体与物体空间关系（他我中心定向）来在环境中定位自己。自我中心定向涉及依赖观察者位置的线索（即左右、前后），而他我中心定向则通过使用地标的环境特征来维持。因此，他我中心定向涉及短期记忆，而自我中心定向则不涉及。本文介绍了一项专门设计用于评估自我中心空间认知的虚拟现实导航（VRN）实验。该设计旨在最小化自然主义VR环境中空间线索或地标的影响对人类导航的作用。为该研究设计的VRN实验，称为虚拟房屋，是一个对称的三层立方体建筑，每层每侧有3扇窗户，建筑每侧有一个入口。在每次试验中，一个窗户被伪随机序列标记为目标。从室外视角向参与者展示标记的窗户。任务是通过建筑物中最短的路径到达目标窗户。该实验需要进行2组，每组8次试验以涵盖所有可能性。参与者的表现误差通过他们的行进距离轨迹与最短自然距离（使用VR引擎计算）之间的差异来衡量，在每次试验中按最短距离进行归一化。52名认知健康的成年人参与了该研究。结果表明在16次试验中没有学习效应，这意味着该实验不依赖短期记忆。此外，受试者的归一化表现误差随年龄几乎呈线性增加，这意味着自我中心空间认知能力随年龄下降。
