The purpose of this study is to investigate the effects of training of the non-involved upper limb on the motor ability of the hemiparetic upper limb in patients with unilateral sensorimotor motor impairments. The secondary aims are to investigate the relation between lesion characteristics and responsiveness to the treatment, and to study which brain structures are activated during a task trained in the treatment.

40 post-stroke subjects will be recruited from neurological rehabilitation department at Loewenstein Hospital Rehabilitation Center, Raanana, Israel. Hospitalized patients, discharged patients from this department and patients that visit the clinic at this department will be recruited. The patients will be randomly assigned to one of two groups: 1) Experimental group. 2) Control group. The 3 weeks intervention (15 sessions) in each group will include daily 30 minutes of structured training followed by 30 minutes of standard physical therapy exercises of the upper limbs (conventional sensorimotor active and passive exercises of the upper limb). During the structured training in the experimental group, patient's hand will be positioned in a specialized motion control apparatus (Rehabit-Tec System). This device consists of a forearm and wrist rest, and the fingers of each hand are individually strapped to the device with palms facing down. The finger is connected to a piston that moves a plunger on a potentiometer according to the degree of flexion. A control module reads the location of every potentiometer on each finger of the 'moving' hand and powers motors that push/pull the corresponding other hand finger to equalize the potentiometer's position. Each finger channel is independent and acts as a stand-alone control circuit. The device restricts voluntary movement of the 'other' hand fingers since only 'moving' hand finger movement activates the motors. Thus when the hands are strapped to the device, and the device is turned on, voluntary hand finger movement results in passive yoking of the corresponding 'other' hand fingers. The patient can't see his real hands and the device. Visual feedback of virtual hands are provided through a VR headset used for 3D gaming. The patient will wear motion-sensing gloves (5DT Data Glove Ultra) that allow online monitoring of individual finger flexure in each hand. The training setup contains a head-mounted specialized 3D camera (PLAYSTATION Eye digital camera device) to provide online visual feedback of the real environment. These devices allow detection of the patient's real hand movements and translate them by customized software to the virtual hand movements presented on the screen. The virtual hands are embedded in a specific location in space and are presented only when the patient look down towards his real hand. In this group, the patients will move their non-involved hand, and will receive a visual feedback of a moving virtual hand in the non-involved side. A passive movement (providing kinesthetic feedback) of the non-involved hand will occur simultaneously by the device. For example: in a left hemiparetic patient, a movement of the right hand will result in observation of the left virtual hand and a passive movement of the left hand. The structured pre-training of the control group will be similar except that the patient will not receive a visual and kinesthetic feedback (passive movement) of the hemiparetic hand. The subjects will be asked to move their non-involved fingers as accurately and fast as they can.

Before and after each intervention the following tests will be used: Fugl-Meyer Assessment and box and blocks test as motor clinical assessment tests of the upper limb, and the ability to move the fingers (using the aforementioned gloves, For example: amount of flexion). Normalized lesion data will be computed using the ABLE module within MEDx software (Medical Numerics). Only before the intervention, sensory ability will be characterized using: Fugl-Meyer assessment of sensorimotor function (Sensory section) and Semmes-Weinstein filaments.

In Cases with specific theoretical interest, the subjects will be offered to be examined also in a fMRI/DTI using Magnetic Resonance Imaging (MRI) System - Siemens Prisma 3T scanner. The task during the imaging will be as during the structured pre-training.

Differences between groups will be done using t-tests or Mann-Whitney, depending on distribution of normality with corrections for multiple comparisons. Correlations or voxel-based lesion symptom mapping (VLSM) analyses will be conducted to investigate the relation between lesion extent and location and responsiveness to the treatments.

Inclusion Criteria:

• Having unilateral motor and/or sensory impairment of the upper limb due to neurological disorder
• Able to provide informed consent

Exclusion Criteria:

• For patients who will be asked to participate in the fMRI/DTI examination only: having pacemakers and metal body