Stroke is a main cause of disability worldwide. It is characterized by motor and cognitive impairments leading to activity limitations and participation restrictions. Despite improvement in mortality and morbidity following stroke, stroke survivors need access to effective rehabilitation services. Management approach for stroke survivors is classical physical and rehabilitation medicine (PRM) interventions. Nowadays, with the advancement of technology, robotic rehabilitation systems have taken its place among the classical physical and rehabilitation medicine applications. Robotic systems for upper and lower extremity help to improve neuroplasticity by repetitive task-specific activities. Upper extremity robotic systems can be either exoskeleton or end-effector according to their mechanical designs. Exoskeleton robotic systems attach to both proximal and distal segments of the upper extremity (shoulder, elbow, forearm, wrist, fingers) and provide antigravity weight support allowing actuated axes of movements of the upper extremity joints. The system allows the reinforcement and facilitation of movements by means of visual feedback with a 3-dimensional virtual environment in which the patient is asked to perform various tasks by playing computer games. It is reported that robotic therapy can be used complementary to other rehabilitation methods. The neurophysiological effects of robotic therapy were shown to act in the brain, particularly on the primary motor cortex, putamen and capsula interna. The effect on motor recovery was associated with common, synchronous activity involving the corticospinal system. It has been reported by a recent Cochrane review that robot-assisted arm training improves arm function, arm muscle strength and activities of daily living in stroke survivors. A few studies have also reported positive effects on cognitive abilities. However, there are no controlled studies in the literature investigating the effects of robot-assisted upper limb training on cognitive functions following stroke. The aim of this study was to investigate the effects of robot-assisted upper extremity training, applied in addition to the classical PRM program, on cognitive and physical functions after stroke. Primary aim is to investigate the effects on cognitive functions whereas secondary aim is to investigate the effects on upper extremity motor functions and activities of daily living. The investigators hypothesized that adjunctive robotic upper extremity training in addition to classical PRM program would result in better cognitive and physical outcomes compared with the classical PRM program only.

This study is planned as a prospective randomized single blind controlled study. It was approved by the Ethics Committee of Ankara University, Faculty of Medicine and will be conducted in accordance with the Declaration of Helsinki. The study will include 38 stroke survivors who are admitted for rehabilitation at the Department of Physical Medicine and Rehabilitation, Ankara University Medical Faculty. Patients fulfilling the inclusion criteria of the study will randomly be divided into 2 groups as robotic group and control group. Random allocation software (RAS) program will be used to assign the patients to the treatment groups with "block randomization" method and to create the randomization scheme. In this single blinded study, all patients will be evaluated by the same physician (Dr. Merve Dayi). The physician will be blind to patient's assignment. Routine physical and rehabilitation medicine program including physical therapy and exercises, walking and balance training, and occupational therapy to improve activities of daily living will be given to all patients for a total of 6 weeks, 5 days/week, 1 hour/day. Robotic group will additionally receive upper extremity robotic training therapy with Armeo Spring V2.0 exoskeleton (Hocoma Inc.,Zurich, Switzerland) for 6 weeks, 5 days/week, 1 hour/day. The Armeo Spring is an adjustable suspension system for the upper limb that connects to virtual reality, which has settings with several degrees of complexity. The system is an exoskeleton that supports the patient's arm and magnifies any residual active movement of the hemiparetic arm in 3-dimensional space. The patients will receive 30 sessions of robotic therapy under the supervision of physiotherapists, who will adjust the patient's arm and hand in the device and select the virtual reality computer games according to the functional level of the patient's upper limb. All individuals participating in the study will be assessed before and immediately after the treatment and also 3 months at follow-up. Each assessment will include evaluation of cognitive functions, upper extremity motor functions, hand dexterity, upper extremity activities and daily living activities.

Statistical Analysis

As the primary outcome variable, the difference in cognitive functions measured by the Montreal Cognitive Assessment Scale (MoCA) between baseline and post-treatment will be used (Δ MoCA). Group sample sizes of 19 and 19 achieve 80% power to detect a Δ MoCA of 2.4 between robotic and control groups with estimated group standard deviations of 2.5 and 2.5 and with a significance level of 0.05 using a two-sided Mann-Whitney U test. Stratified block randomization will be used to ensure an equal number of subjects in each group. Educational status and age has been chosen as stratification variables. Difference between two groups for continuous variables will be evaluated by Student's t test. Mann-Whitney U test will be used to assess difference between two groups in terms of ordinal variables. The differences in proportions between groups will be compared by using Chi-Square. The repeated-measures ANOVA will be used to test within-group and between-group comparisons for continuous variables and interaction between them. p<0.05 will be considered statistically significant.

Inclusion Criteria:

• Patients within the age range of 18-80 years
• Patients with subacute stroke (post-stroke duration 1 month - 1 year)
• Patients who have cognitive deficits and who can understand the commands to adapt to robotic training (MMSE ≤ 26)
• Patients with first stroke attack
• Patients who have never received robotic rehabilitation therapy
• Patients with upper extremity (UE) Brunnstrom Stage III and above
• Patients with a spasticity level below 3 (MAS 1-5 assessment system) according to the modified Ashworth Scale (MAS) in the affected upper extremity
• Patients who agreed to participate in the study and signed the informed consent form

Exclusion Criteria:

• Patients who cannot understand commands or give informed consent
• Patients without cognitive deficit (MMSE > 26)
• Patients with aphasia
• Patients who do not have sitting balance
• Patients with unilateral neglect
• Patients with unstable systemic medical diseases that may prevent the patient to receive robotic training sessions
• Patients with psychiatric disorders
• Patients with posterior cerebral artery infarction and subarachnoid hemorrhage
• Patients with impaired vision that may affect robotic training
• Patients with a peripheral nerve injury or musculoskeletal disease of the affected upper extremity
• Patients who have pain in the affected upper extremity such that this pain might impede the patient to adapt to robotic training
• Patients with involuntary abnormal movements (e.g. dystonia)

Single blind