JMIR Publications

ABSTRACT

Background:

Stroke rehabilitation often requires intensive, repetitive, and engaging interventions to promote motor recovery. Virtual reality–based rehabilitation has emerged as a promising tool to enhance patient motivation and improve functional outcomes. However, the integration of immersive VR systems into routine clinical practice remains limited.

Objective:

To describe the design and iterative development of NeuroRehab VR, a fully immersive, gamified, stroke-specific virtual reality software intended to support upper-limb rehabilitation in patients after stroke.

Methods:

NeuroRehab VR was developed between 2022 and 2024 through a public–private, multidisciplinary, and multicenter collaboration involving neurologists, rehabilitation physicians, physiotherapists, exercise scientists, researchers, engineers, software developers, game designers, and digital artists. The development process followed an agile, patient-centered approach based on the SCRUM methodology and was structured into three phases: preproduction, production, and postproduction. Through iterative development cycles, multidisciplinary meetings, and continuous feedback integration, therapeutic objectives, virtual environments, interaction modalities, gamification strategies, and safety requirements were progressively defined and refined. Preliminary user testing was conducted with seven patients at different stages after stroke through five iterative sessions. Patient feedback, clinical observation, and technical assessment were used to refine usability, task complexity, hand-tracking accuracy, calibration procedures, and interaction design.

Results:

The development process resulted in NeuroRehab VR, a fully immersive software implemented on the Meta Quest 3 headset and based on controller-free hand tracking. Three virtual environments were created: nature, home, and science fiction. Each environment included five therapeutic activities, resulting in 15 gamified tasks targeting fine motor skills, gross motor skills, balance, movement speed, movement sequencing, and cognitive engagement. Iterative testing led to several refinements, including seated task execution to reduce fall risk, alignment between real and virtual tables, simplification or replacement of overly demanding activities, improved calibration procedures, auditory and visual instructions, and the incorporation of a virtual assistant to facilitate user familiarization. A clinician-facing platform was also developed to register anonymized patients, monitor completed sessions, and display performance across therapeutic domains.

Conclusions:

NeuroRehab VR is a stroke-specific immersive virtual reality platform developed through a SCRUM-based, multidisciplinary, and patient-informed process. Preliminary testing supported refinement of its usability, safety, and therapeutic applicability, although clinical effectiveness has not yet been established. Further evaluation in controlled clinical studies is required to determine its effects on disability, upper-limb function, adherence, and quality of life after stroke.