VR and AR rehab platforms exploit the dose-response advantage of immersive engagement: patients tolerate longer sessions and produce more repetitions than they do under conventional therapy alone. The standards stack here is younger and more fragmented than for robotics — IEEE Std 3079-2020 covers HMD-side cybersickness, IEC 80601-2-78 carries over for any robotic component, and the platform is otherwise governed by IEC 62366-1 usability engineering and IEC 62304 software-life-cycle.
Tethered HMDs (Meta Quest Pro, Varjo XR-3, HTC Vive Focus 3 / XR Elite, PlayStation VR2 in research-only deployments) and standalone HMDs each carry distinct hygiene + cleaning + cross-contamination protocols. Per-patient eye-mask liners and face cushions are typical; hospital-grade disinfection per Spaulding-classification semicritical (eye contact) or noncritical (forehead contact) sets the cleaning cadence. Position-tracking accuracy and refresh rate (≥90 Hz, ideally 120 Hz) directly bound cybersickness risk.
The application separates three concerns: (a) the rendering pipeline (OpenXR per Khronos for portability across runtimes); (b) the therapy-protocol engine (per-protocol task graph + per-trial scoring); (c) the session-recording surface that feeds the §A.3 sensor-stream API. Game engines (Unity, Unreal, Godot, A-Frame for WebXR) handle (a). Custom code carries (b) and (c). The clinical-data path runs through the host facility's HIPAA + GDPR + per-jurisdiction privacy envelope.
IEEE Std 3079-2020 prescribes the measurement methodology — Simulator Sickness Questionnaire (SSQ) per Kennedy + Lane + Berbaum + Lilienthal 1993, the four-component Misery Scale, and physiological proxies (HRV, skin conductance, postural sway). Mitigation: snap-turn locomotion, vignette-on-motion, FOV-restriction on translational acceleration, comfortable seated mode, and short-duration tasks (5–10 min) early in the patient's tolerance build-up. The clinical protocol explicitly schedules breaks.
Per Lohse + Hilderman + Cheung + Tatla + Van der Loos 2014, gamified rehab improves engagement but does not necessarily improve outcomes; the design must preserve the therapeutic dose. Self-Determination Theory (Deci + Ryan 1985) frames the autonomy + competence + relatedness drivers. Difficulty adapts trial-by-trial against patient performance; over-easy trials waste dose, over-hard trials elicit avoidance behaviour.
Per WCAG 2.2 + EN 301 549 + ADA Title III, the platform supports alternate input modalities (eye-gaze, voice, switch-based input), captioning for any audio, colour-blind-friendly palettes, and adjustable visual contrast. Patients with photosensitive epilepsy require the per-protocol photic-stimulation envelope per ISO 9241-391 / WCAG 2.3. The platform's per-patient profile retains the per-patient accessibility configuration so the clinician does not re-key the per-patient settings at every session — this is part of the per-patient privacy + consent envelope and must follow the per-jurisdiction patient-data policy.