The Effect of Optic Flow During Treadmill Walking on the Gait Pattern in People Post-stroke

STUDY DESIGN: An experimental 1-group, single-centre trial will be conducted in which people post-stroke will perform 4 sessions of 20 minutes treadmill walking. The first session serves as a control session in which patients will walk without VR. The order of the OF speed (i.e. matched, two times slower than and two times faster than their walking speed) in the following 3 VR sessions will be randomized.

PROCEDURE: Patients will be tested during 4 sessions of treadmill walking in the research lab of the research group R\&MM at the Vrije Universiteit Brussel.

Prior to the start of each session, some preparations will have to be done with regard to the outcome measures:

• Surface electrodes will be placed bilateral on the M. rectus femoris, M. vastus lateralis, M. semitendinosus (hamstrings medialis), M. tibialis anterior, M. gastrocnemius medialis and M. soleus. Electrode placement will follow the SENIAM guidelines. The skin underlying the electrode will be shaved and cleaned with alcohol to improve electrode-skin contact and reduce impedance.
• Reflective markers will be placed on fixed points of the lower limbs of the patient. Marker placement will follow the Plug-in Gait lower body model (VICON).

Session 1: control session After these preparations, participants will be asked to fill in a questionnaire (the Simulator Sickness Questionnaire). Afterwards, participants will start to walk on the patient-controlled treadmill system. For safety only, participants will walk with an additional safety harness. Participants will first be habituated to walking on the treadmill. During this habituation, we will also check if all measurements (VICON and EMG) are correct and that no errors occur. If all measurements are good, participants will walk for 20 minutes on the treadmill. After these 20 minutes, the treadmill will be stopped and participants will be asked to fill in 2 questionnaires (the Simulator Sickness Questionnaire and the Physical Activity Enjoyment Scale).

Sessions 2 - 3 - 4: VR session After these preparations, participants will be asked to fill in a questionnaire (the Simulator Sickness Questionnaire). Afterwards, participants will start to walk on the patient-controlled treadmill system. For safety only, participants will walk with an additional safety harness. At the start of each VR session, participants will first be habituated to walking on the treadmill without VR. During this habituation, we will also check if all measurements (VICON and EMG) are correct and that no errors occur. If all measurements are good, the treadmill will be stopped so that the VR can be added. Participants will put on the head-mounted display (HMD) 'Oculus Rift' and will start to walk in the baseline virtual environment for 2 minutes. After these 2 minutes the appropriate optic flow speed will be added to the virtual world. Participants will walk for 20 minutes in this virtual environment. After these 20 minutes, the treadmill will be stopped and participants will take off the HMD 'Oculus Rift'. To end the session, participants will be asked to fill in 2 questionnaires (the Simulator Sickness Questionnaire and the Physical Activity Enjoyment Scale).

RANDOMIZATION: The first session will always be the one without VR, i.e. the control session. For the VR sessions (session 2, 3 and 4), the optic flow manipulation (i.e. matched, two times slower, two times faster) will be randomized.

MATERIALS: The virtual reality will be provided with the HMD VR system 'Oculus Rift' (Oculus, LLS, US) and will assure an immersive virtual environment. The VR will be used in combination with a treadmill system constructed at the Vrije Universiteit Brussel (Research group R\&MM). The patient-controlled haptic treadmill prototype is a treadmill where the speed is controlled by the natural inclination of the patient while trying to move forward or stand still. The treadmill has 6 axis force/torque sensors connected to a human user through a ball joint that support a harness attached to the patient's pelvis. The treadmill's behavior can be controlled in real time to adapt the walking dynamics of the patient. This enables the treadmill to follow the walking speed of the patient.

STATISTICAL ANALYSIS: The effect optic flow manipulation has on the spatiotemporal gait parameters, kinematics and muscle activity will be compared with walking without VR. In a first stage, the data will be visualized using LO(W)ESS smoothing (locally weighted scatterplot smoothing) to explore the observed effects over time (per condition and outcome), allowing for flexibility using this quasi-nonparametric approach. Next, relevant values expressing onset, magnitude and duration of the effect will be extracted:

• Onset: time point(s) at which the minimal clinically important difference (MCID) is exceeded. In case MCID is unknown, a 10% threshold will be used
• Magnitude: magnitude of the maximum (or maxima) and time point(s) at which the maximum is reached
• Duration: time between the onset and the time point at which the MCID (or 10% threshold) is no longer exceeded

These values will be compared between conditions in a one-way repeated measures ANOVA (factor "condition"). If relevant (based on the exploratory analyses), additional in-depth statistics, such as functional data analysis, will be performed under the guidance of the Department of Statistics and Data Analysis of our university.