Virtual reality driving and brain injury in the clinic

尽管长期以来的文献表明，脑损伤(BI)后驾驶能力的变化--对于这些差异与驾驶失误风险增加或BI后长期驾驶结果的预测之间的关系，人们知之甚少。然而，众所周知，丧失驾驶特权对功能重新融合、情绪和生活质量产生负面影响--这是因为参与各种生活活动、工作和教育经历的能力降低。 提高我们对如何最好地评估和预测BI后的驾驶表现的理解是双重的挑战。首先，需要新的评估方法，能够提供客观、详细和可重复的驱动力绩效衡量标准。目前的临床黄金标准-方向盘后(BTW)驾驶评估过度依赖主观观察，缺乏标准化，只评估基本的驾驶技能(由于安全限制)，并产生粗略的表现衡量标准(即通过/不通过)。其次，缺乏对BI患者实际重返驾驶行为的后续研究。虽然已经有一些证据表明发生车祸的风险更大(通常分为是/否)，但这些研究严重依赖于自我报告的数据，几乎没有提供关于司机行为和/或修改、风险参与、引发车祸的行为或驾驶模式的数据。 这项拟议的研究旨在解决这些限制，并采用已建立的虚拟现实驾驶模拟器(VRDS)，该模拟器可输出目前临床方法学无法获得的新颖驾驶性能指标。VRDS生成可以区分临床人群的详细指标。具体地说，这项研究将把VRDS整合到现有的临床驾驶评估计划中，并在返回驾驶的整个过程中(例如，从评估到后续)评估100名BI患者和健康对照组的样本。所有参与者都将接受当前临床方案和VRDS的评估。随后将进行24个月的跟踪研究，包括一种创新的三平台方法(车载视频监控、基于网络的自我报告和驾驶记录)，以量化回归的驾驶行为。收集的数据将用于应用传统(回归模型)和新的(机器学习模型)分析技术，以生成相关结果变量(即风险参与、与碰撞相关的错误)的预测模型，这些模型可用于为量身定制的司机干预和再培训提供信息。

Despite the long-standing literature that has demonstrated changes in driving capacity following brain injury (BI) – little is known about the relationship of these differences and increased risk for driver error or the prediction of long term driving outcome after BI. However, it is well-established that the loss of the driving privilege negatively impacts functional re-integration, mood and quality of life – resulting from the reduced ability to participate in various life activities, work and educational experiences. The challenge to increasing our understanding of how to best assess and predict driving performance after BI is two-fold. First, there is a need for novel assessment methodologies that can provide objective, detailed and repeatable metrics of driving performance. The current clinical gold standard – the behind the wheel (BTW) driving assessment, is over-dependent on subjective observations, lacks standardization, assesses only basic driving skills (due to safety limitation) and generates gross measures of performance (i.e., Pass/Fail). Second, there is a lack of follow-up studies that examine actual return to driving behaviors among individuals with BI. While some evidence for greater risk of crash involvement (often dichotomized as Yes/No) has been reported, these studies have relied heavily on self-reported data and offer little to no data about driver behaviors and/or modifications, risk-involvement, crash causing-behaviors or driving patterns. The proposed study aims to address these limitations and employs an established virtual reality driving simulator (VRDS) that outputs novel driving performance metrics that are currently not available thru clinical methodology. The VRDS generates detailed metrics that can differentiate between clinical populations. Specifically, the study will integrate VRDS into an existing clinical driving assessment program and evaluate 100 individuals with BI across the process of returning to drive (e.g., from assessment to follow- up) and a sample of healthy controls. All participants will be assessed with both current clinical protocols and VRDS. This will be followed by a 24 month follow-up study including an innovative, 3-platform approach (in- car video-monitoring, web-based self-report and driving records) to quantifying returned to driving behaviors. The data collected will be used to apply both traditional (Regression Models) and novel (Machine-Leaning Models) analytical techniques to generate predictive models of relevant outcome variables (i.e., risk involvement, crash-relevant errors) that can be used to inform tailored driver interventions and retraining.