VRGaitAnalytics: Visualizing Dual Task Cost for VR Gait Assessment

Abstract

Among its many promising applications, Virtual Reality (VR) can simulate diverse real-life scenarios and therefore help experimenters assess individuals’ gait performance (i.e., walking) under controlled functional contexts. VR-based gait assessment may provide low-risk, reproducible and controlled virtual environments, enabling experimenters to investigate underlying causes for imbalance by manipulating experimental conditions such as multi-sensory loads, mental processing loads (cognitive load), and/or motor tasks. We present a low-cost novel VR gait assessment system that simulates virtual obstacles, visual, auditory, and cognitive loads while using motion tracking to assess participants’ walking performance. The system utilizes in-situ spatial visualization for trial playback and instantaneous outcome measures which enable experimenters and participants to observe and interpret their performance. The trial playback can visualize any moment in the trial with embodied graphic segments including the head, waist, and feet. It can also replay two trials at the same time frame for trial-to-trial comparison, which helps visualize the impact of different experimental conditions. The outcome measures, i.e., the metrics related to walking performance, are calculated in real-time and displayed as data graphs in VR. The system can help experimenters get specific gait information on balance performance beyond a typical clinical gait test, making it clinically relevant and potentially applicable to gait rehabilitation. We conducted a feasibility study with physical therapy students, research graduate students, and licensed physical therapists. They evaluated the system and provided feedback on the outcome measures, the spatial visualizations, and the potential use of the system in the clinic. The study results indicate that the system was feasible for gait assessment, and the immediate spatial visualization features were seen as clinically relevant and useful. Limitations and considerations for future work are discussed.