Validation of a Virtual Reality Floor Maze Test to Detect Early Signs of Cognitive Impairment

项目概要/摘要 早期诊断和识别阿尔茨海默病（AD）的预测因子至关重要，因为它允许干预 在神经元损失最小的时候采取预防和治疗策略。空间导航是一个复杂的 以及在脑部疾病早期受损的多成分技能， 为AD的未来临床进展提供相关、敏感和特异的标志物，甚至在其临床前阶段。活性 空间导航评估允许识别与认知下降相关的步态改变， 通过增加行走的姿势要求所产生的双重任务效应来放大测试的难度。 这项研究将验证第一个运动认知筛选仪器能够提取数字标记的形式 有患AD风险的人的导航和步态表现。我们建议一个完全沉浸式的 虚拟现实（VR）导航测试相对于经典测试具有许多优势，因为它允许 根据具体需要对环境特征的操纵。我们将使用VR版本的地板迷宫 测试（VR-FMT）以创建具有首选复杂度的虚拟迷宫，并在商业VR中显示它们 耳机认知正常的成年人，处于AD的低风险和高风险中，以及患有轻度遗忘症的受试者 认知障碍将完成两次访视。在第一次访问中，将进行一系列神经心理学测试。 在第二阶段，参与者将在VR-FMT内执行多个导航。两种视觉表现 （远景和环境），和两个探索类型（真实的步行和与操纵杆）将进行测试。步态会 使用VR追踪器和动作捕捉系统进行记录。功能近红外光谱（fNIRS）将 用于测量静息状态下的大脑连接。在目标1中，我们将验证步态测量和 由VR-FMT操作。我们假设便携式追踪器将显示出相当的有效性 在测量步态相比，光学运动捕捉系统和参与者将显示不同的 在步行和原地版本以及环境和远景空间中的导航性能 VR-FMT的版本。在目标2中，我们将研究VR-FMT作为区分各种 认知障碍的程度。我们假设，导航性能，而执行活动版本 VR-FMT在环境空间中表现出优越的上级区分能力。然后，一台机器 学习算法将用于提取最重要的特征并对参与者进行分类。我们假设 步态和导航性能都将增加分类器的灵敏度。在目标3中，我们将探索 VR-FMT、神经心理学测试和大脑连接之间的关系。我们 假设较低的导航性能将与改变的大脑连接和较低的导航性能相关， 心理速度记忆力和执行功能得分这项研究的结果将为以下方面奠定基础： 进一步的纵向研究旨在预测累积的AD生物标志物。长期目标是 开发一种准确、低成本、用户友好和便携的系统，可用于预测认知能力下降。

PROJECT SUMMARY/ABSTRACT Early diagnosis and identification of predictors for Alzheimer’s Disease (AD) is crucial as it allows intervention with prevention and treatment strategies when neuronal loss is at its minimum. Spatial navigation is a complex and multi-component skill that gets impaired early in the course of brain diseases and may be considered a relevant, sensitive and specific marker for future clinical progress of AD, even in its preclinical stage. Active spatial navigation assessments allow to identify modifications of gait associated with cognitive decline, and to amplify the difficulty of the test through a dual task effect created by the increased postural demands of walking. This study will validate the first motor-cognitive screening instrument able to extract digital markers in the form of navigational and gait performances in people at risk of developing AD. We propose that a fully immersive Virtual Reality (VR) navigational test has numerous advantages with respect to classical tests because it allows the manipulation of environmental features based on specific needs. We will use a VR version of the Floor Maze Test (VR-FMT) to create virtual mazes with preferred complexity and display them within a commercial VR headset. Cognitively normal adults at low risk and higher risk of developing AD and subjects with amnesic mild cognitive impairment will complete two visits. In the first visit, a battery of neuropsychological tests will be taken. In the second, participants will perform multiple navigations inside the VR-FMT. Two visuals representations (vista and environmental), and two explorations types (real walking and with a joystick) will be tested. Gait will be recorded using VR trackers and a motion capture system. Functional near-infrared spectroscopy (fNIRS) will be used to measure resting-state brain connectivity. In Aim 1 we will validate the gait measures and the manipulations operated by the VR-FMT. We hypothesize that the portable trackers will show comparable validity in measuring gait compared to the optical motion capture system and that participants will show different navigation performance in the walking and the in-place versions and in the environmental and vista spaces versions of the VR-FMT. In Aim 2 we will investigate the ability of the VR-FMT as a test to differentiate the various levels of cognitive impairment. We hypothesize that navigation performance while performing the active version of the VR-FMT in the environmental space would show superior ability to distinguish the groups. Then, a machine learning algorithm will be used to extract the most significant features and classify participants. We hypothesize that both gait and navigation performances would increase the sensitivity of the classifier. In Aim 3 we will explore associations between performance in the VR-FMT, neuropsychological tests, and brain connectivity. We hypothesize that lower navigation performance will be associated with altered brain connectivity and lower psychomotor speed, memory and executive function scores. Findings from this research will set the stage for further longitudinal studies which will be aimed at predicting accumulating AD biomarkers. The long-term goal is to develop an accurate, low-cost, user-friendly, and portable system that can be used to predict cognitive decline.