Development of a Virtual Reality-Based Balance Training System for Stroke Rehabilitation

Thesis Abstract

Stroke remains a leading cause of long-term disability worldwide, with balance impairment being a predominant challenge in post-stroke rehabilitation that significantly affects patients' independence and quality of life. Traditional balance training methods often face limitations related to engagement, motivation, and real-time feedback, which may influence adherence and rehabilitation outcomes. In response to these challenges, this study aims to develop and evaluate a Virtual Reality (VR)-based balance training system tailored for stroke survivors. The specific objectives include designing an interactive VR environment that simulates dynamic balance tasks, assessing the system’s usability and efficacy, and comparing its effectiveness against conventional balance training methods. The research adopts a mixed-methods approach comprising both qualitative and quantitative components. A quasi-experimental design will be employed involving a sample of 60 post-stroke patients aged 45-75 years, recruited from physiotherapy clinics within the metropolitan healthcare network. Participants will be randomly assigned to an experimental group receiving VR-based balance training and a control group undergoing traditional physiotherapy-based balance exercises, each with 30 participants. Data collection will involve pre- and post-intervention assessments using standardized measures such as the Berg Balance Scale, Timed Up and Go (TUG) test, and the System Usability Scale (SUS) for evaluating system acceptability. Additionally, semi-structured interviews will explore user experiences to complement quantitative findings. Validity and reliability of the measurement instruments will be ensured through pilot testing and adherence to established psychometric properties. The intervention will span over six weeks, with three training sessions per week. Quantitative data will be analyzed using repeated-measures ANOVA to evaluate changes in balance outcomes over time, supplemented with independent t-tests to compare post-intervention differences between groups. Thematic analysis will be applied to qualitative data, identifying key themes related to user engagement, perceived benefits, and barriers to system use. It is anticipated that the VR-based system will significantly improve balance performance in post-stroke patients compared to conventional therapy, indicated by higher gains on the Berg Balance Scale and faster TUG times. The system's engaging and immersive design is expected to enhance motivation, leading to higher adherence rates. The study also aims to demonstrate high usability scores and positive user experiences, supporting the system’s practicality for clinical integration. This research contributes to the body of knowledge by providing empirical evidence on the efficacy of VR technology in stroke rehabilitation, particularly addressing the gap regarding balance-specific interventions. It also presents a scalable, cost-effective, and user-centered approach that integrates advanced ICT solutions into neurorehabilitation frameworks. The findings will inform clinicians and technologists about how VR can complement existing therapeutic protocols, optimize patient outcomes, and foster innovative rehabilitation practices. The study concludes that VR-based balance training systems are a promising adjunct in stroke rehabilitation, offering improved engagement and functional recovery. Recommendations include further large-scale studies to validate long-term benefits, exploring customization features for individual patient needs, and developing guidelines for integration into routine clinical practice. Future research could expand into exploring augmented reality interfaces and remote, tele-rehabilitation applications, thus broadening access and optimizing stroke recovery trajectories across diverse settings.