Balance Wood, a Bilateral Ankle-Controlled Serious Game for Healthy Adults: 2-Stage Cross-Sectional Development and...

Ankle mobility and control contribute to balance and everyday movement throughout adulthood, yet are usually assessed through repetitive, abstract tasks that reveal little about movement production. Serious games can make repeated tasks engaging, but most report only end point scores rather than continuous gameplay signals, and few establish system usability. Developing and evaluating such a system in healthy adults is a necessary first step.

This study aimed to develop Balance Wood, a bilateral ankle-controlled serious game in which seated players tilt a device through ankle flexion-extension to control on-screen play, to describe the system and the signals it records, and to evaluate its usability, safety, and user experience in healthy adults.

In this two-stage cross-sectional development and usability study, healthy adults without lower-limb injury or balance-affecting neurological or vestibular disorders completed single-session repeated-measures procedures. In stage 1, a total of 40 adults-20 older (mean age 73.0, SD 5.9 y; 10 women) and 20 younger (mean age 28.1, SD 3.2 y; 4 women) adults-completed 3 fixed-order difficulty levels in 1 session (exploratory signal analysis). In stage 2, a total of 18 younger adults (mean age 24.6, SD 3.9 y; 5 women) completed the levels in counterbalanced order. Usability was assessed with the System Usability Scale (0-100); use impressions and consciously used body parts with Likert and open-ended items; and device tilt-angle accuracy against a digital goniometer.

The mean System Usability Scale score was 81.5 (SD 11.2; 95% CI 76.0-87.1; Cronbach α=.74), above the benchmark of 68, indicating good perceived usability. Mean ratings were 2.11 (SD 1.32) for safety concern (95% CI 1.45-2.77) and 1.67 (SD 1.03) for pain (95% CI 1.15-2.18). Ankle use was highest at 4.94 (SD 0.24), exceeding knee 2.50 (SD 1.29), hip 2.00 (SD 1.24), and trunk 2.17 (SD 1.15) by Wilcoxon signed-rank tests (P<.001). Device readings agreed closely with the goniometer, with a mean absolute error of 0.66°, an intraclass correlation coefficient of 0.999, and 95% limits of agreement of -0.68 to 1.70°. In stage 1, all measures were computable for the 40 participants, with typical amplitudes of 20.5° to 27.2° and stable speed in both groups. In stage 2, repeated-measures analyses (α=.05) did not detect statistically significant differences across difficulty conditions or presentation positions (all P>.13).

In healthy adults, Balance Wood was usable, well-accepted, low-risk, and provided accurate device-level tilt measurement with predominantly ankle-based control while continuously recording gameplay signals. It is a low-cost, bilateral, ankle-specific control interface with continuous signal capture into 1 easy-to-use device. Its contribution is a reproducible, usability-tested platform with characterized device accuracy, and formally defined candidate measures awaiting validation. Once validated, it could support home- or community-based recording; in the near term, it is a research tool.