The Eyes Have It: Visual Feedback Methods to Make Walking in Immersive Virtual Reality More Accessible for People With Mobility Impairments While Utilizing Head-Mounted Displays
M. Rasel Mahmud, Alberto Cordova, John Quarles · 2023 · Proceedings of the 25th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS 2023) · doi:10.1145/3597638.3608406
Summary
This paper addresses a critical accessibility barrier in VR: head-mounted displays (HMDs) obstruct peripheral vision of the real world, causing gait disturbances that are particularly dangerous for people with mobility impairments who rely on visual cues for balance. The researchers investigated three types of visual feedback overlaid in VR to improve walking performance: spatial feedback (a crosshair texture fixed to the front wall, stationary relative to the user), static feedback (the same texture but moving with the user's head direction, like a HUD reticle), and rhythmic feedback (identical to static but flashing at one-second intervals). The study involved 50 participants — 25 with mobility impairments due to multiple sclerosis (MS) and 25 without — who completed timed walking tasks on a GAITRite pressure-sensitive walkway system in both real-world and VR conditions. The HTC Vive Pro Eye HMD was used, and participants wore safety harnesses. Twelve gait metrics were measured including walking velocity, cadence, step length, stride length, step time, cycle time, and swing time. The study used a rigorous 2x5 mixed-model ANOVA design with counterbalanced conditions and Bonferroni corrections.
Key findings
VR without visual feedback significantly reduced walking velocity for both groups compared to real-world walking (p < .001), confirming that HMDs alone cause gait disturbances. For participants with MS, static and rhythmic visual feedback both significantly improved three key gait metrics — walking velocity, step length, and stride length — compared to the VR baseline without feedback (p < .001, Cohen's d = 0.71 and 0.75 respectively, medium effect sizes). Spatial visual feedback had no significant effect. Static and rhythmic feedback both significantly outperformed spatial feedback (p < .001), likely because spatial feedback was fixed to the wall and didn't move with the user's view, providing insufficient visual input. There was no significant difference between static and rhythmic conditions. Crucially, for participants without mobility impairments, none of the visual feedback methods produced significant improvements — the feedback specifically benefited those with MI, supporting the hypothesis that people with mobility impairments rely more heavily on visual cues than those without. No significant cybersickness effects were found for either group, and the visual feedback helped participants walk straighter by providing directional orientation.
Relevance
This research directly addresses VR accessibility for the estimated 12% of adults globally who have mobility difficulties. As VR becomes more common in rehabilitation, fitness, education, and entertainment, ensuring that people with mobility impairments can safely walk while wearing HMDs is essential. The practical finding that simple visual overlays (static crosshair textures that move with the user's view) can significantly improve gait performance provides an immediately implementable solution for VR developers — these feedback methods require minimal computational overhead and could be added as accessibility options in any VR application. The differential effect between MI and non-MI groups suggests that these visual aids don't distract typical users while meaningfully helping those who need them. For rehabilitation professionals, the results support using immersive HMD-based VR rather than less effective projector-based systems for gait therapy.
Tags: virtual reality · mobility impairment · multiple sclerosis · gait · visual feedback · head-mounted display · accessibility · balance