To determine the added value of integrating virtual reality therapy (VRT) based balance training into conventional therapy (CT) for improving balance and gait in subacute stroke patients, and to evaluate the usability and patient satisfaction associated with the VRT based intervention, stroke patients will be recruited in the Neurorehabilitation Department of Erasme University Hospital.

Patients will be randomized to a group performing only CT (CT group) and a group in which 20% of the daily CT was replaced by VRT involving standing balance exercises performed with an immersive headset (VRT group). Functional ambulation classification (FAC), gait speed and step length on 10 meters walking test (10MWT), walking distance on 2 minutes walking test (2MWT) and Berg Balance Scale (BBS) will be assessed at baseline (T0) and post-intervention (T1). Sensory reweighting in the VRT group will also be evaluated through the Sensory Organization Test (SOT) at T0 and T1. Usability of the VR system and patient satisfaction associated with virtual reality will be assessed at T1 in the VRT group through a questionary.

Outcomes :

Analyse the effects of a virtual reality balance rehabilitation programme added to conventional therapy on the balance and walking ability of stroke patients.

Analyse feedback from patients who have experienced virtual reality in terms of the usability of the system and patient satisfaction associated with virtual reality.

Participants :

Stroke patients will be recruited in the Neurorehabilitation Department of Erasme University Hospital. Inclusion criteria will be

• Hemorrhagic or ischemic stroke within the past 6 months,
• Age of more than 18 years old,
• Ability to stand and walk with or without assistance
• Mini-Mental State Examination score of more than 15/30
• Ability to understand instructions,
• No visual deficits hindering VR training and vi) a health status allowing for rehabilitation Exclusion criteria will be another orthopedic or neurological disease altering the function of the paretic lower limb, balance or gait.

Study design This study will be a pragmatic, randomized controlled trial conducted in routine clinical settings. Participants will be enrolled during their inpatient stay in the neurorehabilitation unit. Following enrolment, they will be randomly allocated to either the CT group (conventional therapy alone) or VRT group (CT + VRT program) using a computer-generated randomization procedure.

• Control group : conventional therapy alone will focus on motor rehabilitation, individualized according to their specific functional needs (exercises for individual deficits and mainly comprised sit-to-stand transfers, balance exercises, walking independence exercises, ergocycle endurance exercises, and dual-task exercises)
• VRT group : Conventional therapy + virtual reality therapy. Conventional therapy will be partially replaced by VRT (20% of the total weekly rehabilitation time over a 2-week period corresponding in five 20-minute segments each week, resulting in a total of 10 VRT sessions) in such a way that the overall daily duration of motor rehabilitation will be equivalent between the VRT and CT groups.

Virtual reality therapy The virtual reality system will consist of a fixed computer running Virtualis software (Virtualis, France) an HMD (HTC Vive, HTC, Taiwan) with a resolution of 2,160 × 1,200 pixels (Full HD), a horizontal field of view of 110◦ and a frame rate of 90 Hz, two portable controllers (Vive Controller, HTC, Taiwan) and two static force platforms designed to measure center of pressure (CoP) data synchronized with the immersive virtual environments. Each session will last 20 minutes and will be performed under therapist supervision The VR training will consist of different types of exercises: 1/ weight transfer exercises (e.g. skiing) 2/ scrolling exercises (e.g. scrolling in a tunnel); 3/ dual-task exercises (balance + fun upper limb exercises) and 4/ stability limit exercises.

Assessment Patients will be evaluated at two time points: at baseline (T0) and post-intervention (T1, after the two weeks intervention) using validated outcome measures. Functional Ambulation Classification (FAC) will be used to quantify the level of human assistance required for ambulation. Walking capacity will be assessed through the Two-Minute Walk Test (2MWT). Walking speed and stride length will be measured using the 10-meter walk test (10MWT). Postural stability will be evaluated using the Berg Balance Scale (BBS). In addition, participants assigned to the VRT group will undergo a virtual reality-based Sensory Organization Test (VR-SOT) to objectively assess postural control at T0 and T1, and they will be asked to complete a structured satisfaction questionnaire on VR at T1.

Statistical analysis Normality will be verified using the Shapiro-Wilk. For the the primary outcome analyses, a two-way repeated-measures ANOVA will be conducted to examine the group (CT vs. VRT), time (T0-T1) and group x time interaction.

Within the VRT group, changes in VR-SOT scores between T0 and T1 will be analyzed using the Wilcoxon signed-rank test with Holm-adjusted p values. Spearman correlation coefficients will be calculated to assess the relationships between changes in SOT scores and changes in functional outcomes.

Inclusion Criteria:

• Hemorrhagic or ischemic stroke within the past 6 months,
• Minimum age of 18 years old,
• Ability to stand and walk with or without assistance
• Mini-Mental State Examination score of more than 15/30
• Ability to understand instructions,
• No visual deficits hindering VR training and vi) a health status allowing for rehabilitation

Exclusion Criteria:

• other orthopedic or neurological disease altering the function of the paretic lower limb, balance or gait.