JMIR Publications

ABSTRACT

Background:

Impaired motor function and balance deficits increase fall risk in People with Stroke (PwS), potentially leading to injury and reduced physical activity. Perturbation-based balance training (PBT), which exposes individuals to repeated balance disturbances, has demonstrated strong potential for fall prevention but is typically limited to specialized rehabilitation clinics. Extending this approach to home environments could improve accessibility and long-term adherence. Action Observation combined with Motor Simulation (AOMS) has been shown to enhance reactive stepping responses and offer a safe and feasible pathway to translate key elements of PBT to the home environment. Reactive step training, when integrated with exergames, can further enhance motivation and engagement in neurological rehabilitation.

Objective:

To design a home-based exergame that integrates AOMS-based reactive step training and support virtual perturbation-inspired balance training for PwS. In addition, to evaluate the user experience in terms of game experience, usability, and safety perception.

Methods:

Using a user-centered design methodology, we iteratively developed the HEROES exergame prototype for home-based rehabilitation. The exergame, was evaluated as a proof-of-concept during a single supervised session to support home based rehabilitation by training reactive stepping responses. The exergame combines physical exercise with cognitive tasks in a gamified virtual environment where the player needs to imagine and mimic the stepping responses of an avatar in response to virtual perturbations. The prototype was evaluated by PwS (n=8) with a single play-test session conducted in a simulated home environment. Game experience was assessed by the Game Experience Questionnaire (GEQ), usability with the eHealth Usability Benchmarking Instrument (HUBBI) and safety perception by a custom questionnaire, all using 5-point Likert scales. In addition, observations and comments from participants and physiotherapists were collected using a think aloud approach to gain deeper insights.

Results:

Eight PwS participated in the play-test. GEQ scores showed high median ratings for Competence (4.2), Flow (3.9), and Positive Affect (4.7), Tension (1.0), Negative Affect (1.6), and Challenge (2.3) were low. HUBBI results revealed strong Task–Technology Fit (4.3), Interface Design (4.5), and Navigation (4.0), but lower Basic System Performance (2.5), indicating technical issues, such as error in the tracking recognition, that could hinder usability. Safety perception ratings were high: Physical Safety medians were 5.0 (both groups), Trust medians were 4.4 (PwS) and 4.9 (physiotherapist), while Psychological Safety (the feeling of falling while playing the game) was lower (3.7 for both groups).

Conclusions:

The HEROES exergame prototype demonstrates that integrating AOMS principles into a home-based perturbation-inspired balance training approach is feasible. PwS perceived the system as engaging, usable, and physically safe, though improvements in technical robustness and task challenge may be required to sustain long-term engagement. Future work should explore longer-term use and assess clinical effectiveness to determine whether it can improve reactive stepping performance and reduce falls.