Effect of Robotised Gait Training on Dynamic Balance, Symmetry and Push-off in Persons After Stroke (BART)

Rehabilitation robotics is increasingly used because it enables the patients to practice a wide array of movements. Dynamic balance training is essential for gait rehabilitation and robotised devices enhance repeatability, objectivity and precision of such training combined with monitoring and recording of kinematic and kinetic data. The aim of the study is to explore the effect of robot-assisted gait training on dynamic balance, symmetry and take-off in patients after stroke. The investigators will conduct a randomised intervention study where one group will receive visual feedback on gait status and the other group will receive kinetically-assisted training using a robotised device in addition to the visual feedback.

Gait training will start 3 weeks after admission to inpatient rehabilitation. It will last for 3 weeks, 5 times per week, 30 minutes per day. The first few sessions with the robotic device will serve to familiarise the patient with the BART device and training conditions. Further details are described in the Arms and Interventions section.

In each session, the group will undergo standard gait training on BART without pelvic perturbations, i.e., walk a virtual track on the BART without pelvic perturbations (i.e., the pelvic brace of the BART device will be set to follow the patient's motion) for 10 minutes; practise gait symmetry and take-off using visual feedback without pelvic perturbations in two 10-minute sessions.

In each session, the group will undergo robotised gait training with BART with pelvic perturbations, i.e., walk a virtual track on the BART with pelvic perturbations during virtual uphill walk and virtual curved walk for 10 minutes; practise gait symmetry and take-off using visual feedback with pelvic perturbations for 10 minutes; practise dynamic gait balance using pelvic perturbations during treadmill walking for 10 minutes.

The Balance Assisted Robot on Treadmill (BART) enables various types of gait training on treadmill with visual feedback. It interfaces to the pelvis of the walking subject in an actuated and admittance-controlled manner, thus providing transparent haptic interaction with negligible power transfer.

In addition to the standard gait training, the BART will deliver perturbations in the forward/backward and left/right direction.

Assessed using the Functional Gait Assessment (FGA) scale. The scale scores range from 0 to 30, with higher scores indicating less impairment.

Assessed using the Functional Ambulation Categories (FAC) scale. The FAC is a 6-point scale (0-5), with higher category indicating better walking ability.

Assessed using the Motor subscale of the Functional Independence Measure (FIM). FIM Motor subscale scores range from 13 to 91, with higher scores indicating better functional independence.

Assessed using the Timed Up and Go (TUG) test. Longer TUG times indicate worse mobility (and thus higher fall risk).

Assessed using the Four Square Step Test (FSST). The FSST is timed, with longer times indicating worse ability.

Inclusion Criteria: Subacute phase after first ischaemic or haemorrhagic stroke (less than 6 months after stroke); Limb hemiparesis; Ability to walk independently (FAC 6) or with assistance (FAC 5) on flat and uneven surfaces, stairs and slopes; Adequate cardiovascular ability (NYHA 1). Exclusion Criteria: Degenerative process or postoperative condition on lower-limb joints that would hinder gait; Associated neurological disease; Decreased cognitive ability (KPSS < 25).