Virtual reality as an assessment tool in neurorehabilitation: a scoping review of current evidence and future...

BACKGROUND: Virtual reality (VR) technologies are increasingly applied in neurorehabilitation, but most research has focused on their therapeutic applications rather than assessment potential. OBJECTIVE: The aim of this scoping review was to map the evidence on VR-based assessment tools in neurorehabilitation, including types of tools, targeted neurological conditions, technological specifications, and research gaps. METHODS: A scoping review was conducted and reported in accordance with PRISMA-ScR. The protocol was prospectively registered in PROSPERO (CRD420251113260). Searches of PubMed, Scopus, Web of Science, The Cochrane Central Register of Controlled Trials (CENTRAL), IEEE Xplore, and grey literature were performed from inception to July 2025. Eligible studies involved VR for assessment purposes in neurological populations and reported validity, reliability, usability, or assessment performance. Data were charted and synthesised thematically. RESULTS: Sixteen studies met the inclusion criteria. Most adapted established assessments such as the Box and Block Test, Action Research Arm Test, or Virtual Peg Insertion Test into VR formats. Populations included stroke, multiple sclerosis, Parkinson’s disease, cerebral palsy, spinal cord injury, and movement disorders. VR systems ranged from custom-built motion-tracking platforms to commercial head-mounted displays with controllers or optical hand-tracking. Convergent validity ranged from moderate to excellent (r = 0.65–0.99), with optical tracking and mixed reality systems (r > 0.90) generally outperforming controller-based approaches (r ≈ 0.65). Usability was often rated highly but varied across conditions and modalities, with some studies reporting higher usability ratings for physical tasks compared to VR. Key gaps included limited normative datasets, heterogeneous hardware/software, and minimal longitudinal evaluations. CONCLUSIONS: VR-based assessments in neurorehabilitation remain a developing field. Preliminary evidence suggests good validity, usability, and feasibility, with potential to provide richer performance data that augment conventional assessment. However, large-scale, standardised, and condition-specific studies are required before widespread clinical implementation.