Poster W10 - Feasibility of at-home gait rehabilitation using a virtual reality headset

Introduction: Rehabilitation is often necessary to regain mobility, muscular strength, and balance after an injury. This rehabilitation usually involves both in-clinic exercises with a trained physical therapist combined with at-home exercises, such as treadmill walking. Although a strict training regimen is necessary for patients to reach their performance goals, rehabilitation can be a dull and unexciting process for patients exercising at home, with only 36% completing their care plans and achieving maximum results. New and emerging technology such as virtual reality (VR) is showing promise for engaging patients in the rehabilitation process and maximizing patient outcomes. VR enables practical, cost-effective home-based rehabilitation. For example, rehabilitation for gait and balance on a treadmill may be enhanced by adding virtual reality environments that challenge sensorimotor integration and/or simulate ecologically valid environments outside the home. However, virtual reality gait rehabilitation has been limited to large and expensive ecosystems such as the Computer Assisted Rehab Environment (CAREN) that have positive patient outcomes in the clinical environment but are too large, cost several millions, and are too complex for an in-home solution [4]. With the emergence of affordable VR headsets, there is now an opportunity to investigate practical applications of VR for gait rehabilitation that could be feasibly deployed in a patient’s home. Our study aimed to determine the feasibility of VR for in-home gait rehabilitation using consumer-grade hardware.

Materials and

Methods: Nine able-bodied adults (55% female, age: 25±5 years, height: 1.5 – 1.8 meters) wore a Meta Quest 3 headset while walking at 2 mph on a treadmill. Participants were shown one of three custom virtual walking environments (a desert, a forest, or a city) built using the Unity engine. The virtual environments simulated walking on a straight narrow path at a speed matched to the treadmill. Hills and valleys were programmed into the environment to match the treadmill’s change in incline based on preset programs. Each VR environment incorporated different levels of visual stimuli that can be adjusted based on patient needs outlined by their physical therapist. To assess gait performance in real-time, a Logitech C920x web- -camera was configured for marker-less motion capture using an open-source platform called FreeMoCap. We instructed subjects to walk on the treadmill at 2 mph without a virtual environment (i.e. standard treadmill use) for 5 minutes and asked them to rank their engagement on a scale from 1-10. Next, patients were instructed to wear the headset and perform the same exercise for the same duration while immersed in a randomized VR environment and rank their experience on the same scale.

Results, Conclusions, and Discussions: Our results suggest that VR headsets have feasibility for at-home gait rehabilitation. Our system was affordable (total cost <$1K) and required minimal space to deploy (2x1 meter footprint), which makes this rehabilitation approach accessible for many consumers. We also found that the VR environments doubled the average participant’s engagement when compared to unstimulated exercises (Fig. 1), which may help improve adherence to an exercise regimen. Participants also reported the VR environments were fun and did not cause eye strain or motion sickness. VR is an engaging and immersive technology that can encourage participation in exercise [6]. This feasibility study supports a practical application of VR to supplement traditional rehabilitation, and may also support a platform to integrate other consumer-grade hardware into an at-home rehabilitation ecosystem (e.g., wearable sensors, real-time monitoring, coaching, gaming). Future studies should deploy VR environments in clinical studies to get feedback from patients and physical therapists for iteration and improvement.

Acknowledgements (Optional): This design project was supported by the Bioengineering Department at George Mason University