Feasibility of a novel augmented reality application for eliciting altered spatiotemporal gait patterns
Abstract
Use of augmented reality (AR) technology for rehabilitation has drastically increased in recent years. Through providing visual and auditory cues while walking, AR can be used to cue spatiotemporal gait adaptations with applications to clinical populations such as those with Parkinson’s disease. However, given the novelty of the technology it is unknown how cues delivered through AR drive gait adaptations. PURPOSE: To assess the feasibility of visual and auditory cues delivered through AR on altering spatiotemporal gait outcomes in a healthy, young population. METHODS: 20 healthy participants (7 F/13 M; 25.5 + 4 yrs) were recruited to walk for 10 steps in four different cueing conditions using an AR headset: No Cues (NC) (i.e., natural gait), Visual (V), Auditory (A), and Visual + Auditory (VA). Each condition was completed three times in a random order for a total of 12 trials. An Inertial Measurement Unit (IMU) system with integrated footswitches was used to collect spatiotemporal gait data at 200 Hz. A System Usability Survey (SUS) was administered afterwards to determine the usability of our novel application and linear regressions were performed to determine the relationship between reported usability and gait variability. RESULTS: All cueing conditions exhibited a significantly faster cadence (V, A, VA = 0.67, 0.68, 0.68 steps/sec, respectively) compared to NC trials (0.63 steps/sec; p <0.05). Surprisingly, cadence variability was significantly higher for A trials (coefficient of variation (CV) = 0.11) compared to the other three groups (NC, V, AV = 0.1, 0.09. 0.09; p <0.05). V trials exhibited significantly decreased stride lengths (2.13 m) compared to NC (2.3 m; p <0.05). Increased reported system usability was significantly correlated with decreased stance time across A trials (adj R2 = 0.262, β = -0.549, p = 0.012). CONCLUSIONS: Our findings reinforced that certain visual and auditory cues affect gait parameters, albeit in a direction opposite of what was hypothesized (e.g., greater cadence variability with auditory cues). These results provide insight into how healthy populations respond to cues delivered through AR, providing a foundation for future studies to implement AR with clinical populations such as those with Parkinson’s disease. Supported by DoD, US Army (Contract #W911NF2110273)
Gwendolyn Retzinger
Master’s Candidate
Borna Golbarg
Undergraduate Student
Rabbani Nzeza
Master’s Candidate
