Lateropulsion is a deficit in the body orientation with respect to the vertical in the coronal plane, defined by the presence of one of the three signs: lateral body tilt, active pushing from the sound limbs, and resistance to passive corrections. The lateral body tilt is the cardinal sign, the frequency of the 2 other signs increasing with lateropulsion severity (most dramatic forms called pusher syndrome in the past). Lateropulsion is frequent after stroke, and represents the main factor underpinning balance and gait disorders at the subacute phase. After hemisphere stroke lateropulsion is caused by a bias in the internal model of the verticality in the frontal plane, individuals unconsciously aligning their body posture on a tilted verticality representation. Pilot studies suggested the possibility to recalibrate the internal model of verticality, biased by stroke, and to improve individuals' uprightness. The investigators expect that a specific rehabilitation program combining technics devoted to lateropulsion, and comprising exoskeleton (Atalante) assisted balance exercises could help recalibrate the internal model of verticality and alleviate lateropulsion. The primary objective is to test the hypothesis that a 3-week specific lateropulsion rehabilitation program (15 sessions of 30 minutes including exoskeleton and a rehabilitation focused on the vertical body orientation in the frontal plane) improves the visual vertical (VV), the most used test to assess verticality perception.

BACKGROUND To our knowledge, no study has ever assessed the effects of a rehabilitation program specifically devoted to lateropulsion. The investigators expect that such a specific program combining different technics selected on the basis of theoretical rationale, experimental evidence in healthy subjects, or phase I-II randomized clinical studies, and supposed to improve verticality representation or alleviate lateropulsion could improve verticality representation and alleviate lateropulsion. The investigators hypothesize that individuals with moderate to severe lateropulsion after hemisphere stroke could benefit from such a specific lateropulsion rehabilitation program.

OBJECTIVES The primary objective is to test the hypothesis that, in individuals with a first ischemic right hemisphere stroke showing a moderate to severe lateropulsion due to a biased internal model of verticality, a 3-week intensive specific and intensive lateropulsion rehabilitation program (15 physiotherapy sessions of 30 minutes including exoskeleton-assisted lateral balance exercises) improves the visual perception of the verticality as assessed by the Visual Vertical (VV) orientation. This outcome has been selected because it may be repeatedly assessed, which is necessary in a Small-N design Study.

The secondary objectives are to:

• Test the hypothesis that the specific and intensive lateropulsion rehabilitation program also modulates the postural perception of the vertical (PV). This transmodal modulation of verticality perception involving both VV and PV would mean that the internal model of verticality has been recalibrated. This would have a clinical and scientific significance going far beyond a simple change in verticality perception
• Test the hypothesis that the specific and intensive lateropulsion rehabilitation program alleviates lateropulsion;
• Test the hypothesis that the specific and intensive lateropulsion rehabilitation program improves balance and gait abilities without any effect on the motricity of the paretic upper limb;
• Evaluate the participants' perception at the end of this specific rehabilitation program.

HYPOTHESES Considering the main objective, the investigators hypothesize that the specific and intensive lateropulsion rehabilitation program will improve VV orientation of at least 2°, constituting a real change in VV perception.

STUDY DESIGN The study is a monocentric Small-n design study. The experimental plan is a multiple baseline design. This design applies for intervention with slow/delayed changes and no expected washout. Three participants will be included after about the same time post stroke (subacute phase). The VV will be measured daily (except on weekends) for each participant (i.e. 5 VV per week ), and the generalization and control measures measured once each week. All participants will follow a conventional rehabilitation program until they start the intervention.

One participant will start the intervention period 2 weeks after enrolment. Another will start at 3 weeks after enrolment and the last one will start at 3 weeks after enrolment. All participants will be assessed repetitively on the target variables. The sequential introduction of the intervention in 3 participants will allow the visualization of lack of retest effect and of progress unrelated to intervention in the individuals not having the intervention yet.

INTERVENTION

During the baseline and after the 3-week intervention, individuals will follow the conventional rehabilitation program including :

• 2 sessions per day of 30-minutes of "conventional" physiotherapy 5 days a week without exercises specifically devoted to the control of body orientation in frontal plane, or verticality representation
• 1 session of 30-min of occupational therapy 5 days a week
• speech therapy and neuropsychology if necessary.

During the 3-week intervention, individuals will have:

• 1 session per day of 30-min of physiotherapy specifically focused on the body orientation with respect to gravity in the frontal plane, including exoskeleton-assisted balance and gait exercises 3 times a week, 5 days a week,
• 1 session per day of 30-min conventional physiotherapy 5 days a week
• 1 session of 30-min of occupational therapy 5 days a week
• speech therapy and neuropsychology if necessary. So, the intervention phase lasts 3 weeks, and consists in 15 physiotherapy sessions, focused on the active vertical body orientation in the frontal plane and comprising exoskeleton-assisted balance and gait exercises. Participants have the same dose of rehabilitation in each phase (baseline, during intervention and after intervention), which only differ in the nature of the rehabilitation.

During the specific rehabilitation program, the frame of exercises with the Atalante® exoskeleton comprises verticalization: work in standing position, stepping, walking in different directions including backward, with the help of a mirror or without visual information (eyes closed). Some exercises in the exoskeleton will also be performed when the participant is laterally tilted (in a static position, during active functional tasks, eyes open and closed).

The specific physiotherapy rehabilitation targeting lateropulsion will include: verticality awareness, exercises in front of a mirror to facilitate the awareness of body tilt (or with a video record), posturography with biofeedback, trunk inclination, reaching tasks in sitting and standing position, walking with body-weight support, orientation tasks during sensorial ponderation (eyes open/closed, foam or firm surface, cervical extension...) The investigators chose not to include brain stimulation in this program.

PARTICIPANTS Three patients will be included. The study will be explained to eligible individuals who will be free to participate or not. An informed consent form will then be signed by each participant.

To meet the scientific standards of SCEDs, one required design criteria is to include at least three attempts to demonstrate an intervention effect (i.e. 3 participants in our case). The sequential introduction of the intervention in 3 participants will allow the visualization of effect of the intervention on the VV.

RELEVANCE Regaining independent balance and mobility is a major goal in post-stroke rehabilitation. While lateropulsion due to a biased internal model of verticality is one of the most important detrimental factors affecting balance and gait at the post-stroke subacute phase, no specific lateropulsion rehabilitation program has ever been scientifically validated with a high level of evidence. This study will focus on individuals with moderate to severe lateropulsion caused by a biased internal model of verticality due to a right hemisphere stroke which are known to have lesser recovery and longer hospital length of stay. A specific and intensive lateropulsion rehabilitation program dedicated to the active vertical body orientation in the frontal plan could speed up the mobility recovery and shorten the hospitalization period.

Inclusion Criteria:

• Right hemispheric first and unique ischemic stroke between 2 weeks and 6 months old (subacute phase) ;
• Presence of lateropulsion, assessed by the Scale for Contraversive Pushing (2 ≤ SCP < 5)
• Presence of contralesional VV bias (VV > 4°)
• Right handedness defined by a Edinburgh score ≥0,39
• Being hospitalized in a Physical Medicine and Neurological Rehabilitation (PMR) facility.
• Have given a written and informed consent

Exclusion Criteria:

• Medical instability making evaluation impossible,
• Comprehension or cognitive or behavioral disorders, or depressive symptoms that compromise participation in the program
• Postural disorders or body geometry disorders that interfere with balance for a reason that is independant of the stroke
• Morphological contraindications to the use of the Atalante® robot (segment length or joint ranges; height <155cm; weight >100kg),
• Severe spasticity (>3 on the modified Ashworth scale) in the adductors, quadriceps, hamstrings and triceps surae
• Presence of a pressure sore at the areas of contact with the Atalante® robot,
• History of osteoporotic fractures.

Anonymized data that support the findings of this study will be available from the principal investigator (Pr Dominic Pérennou), upon reasonable request, only for authorized research. Their use is subjected to an agreement with the promotor (CHU Grenoble) and the principal investigator of the EXOLAT study

After seminal publication of the results

Anonymized data that support the findings of this study will be available from the principal investigator (Pr Dominic Pérennou), upon reasonable request, only for authorized research. Their use is subjected to an agreement with the promotor (CHU Grenoble) and the principal investigator of the EXOLAT study