Rationale: Approx. 80% of acute stroke patients suffer from acute hemiparesis. Stroke patients have not reached their full potential when they are discharged from hospital. It is proven that extra training opportunities lead to further improvement. To date, new training possibilities, such as robotic techniques for rehabilitation, virtual reality training systems and tele-rehabilitation are being developed to aid in the training of stroke patients.

Objective: To obtain preliminary evidence on the efficacy of an individualised, intensive 6-week technology-assisted training regime, featuring a robotics-based self-adapting arm training system (I-TRAVLE) in a virtual context, focussed on improvement of arm function and arm skill performance in chronic stroke patients with low to moderate proximal (shoulder/arm) muscle strength.

Study design: single arm prospective cohort study. Study population: 16 stroke patients in the chronic phase after their stroke, aged >=18, diagnosed with a central paresis of the arm, having low to moderate proximal (shoulder and arm) muscle strength.

Intervention (if applicable): Haptic feedback of the I-TRAVLE robot either supports or challenges the patient to perform movements of the arm, thereby training motor control and co-ordination of the affected arm. Also strength and endurance of arm muscle use may be trained. The I-TRAVLE based training will last 6 weeks. Each week participants will attend training sessions on 3 days, during which they will train 2x 30 minutes, interspaced by at least half an hour to avoid (general) fatigue and overuse.

Main study parameters/endpoints: Baseline measurements will be performed 3x prior to the start of the intervention, each interspaced by 1 week to assess baseline stability or any fluctuations in baseline values. In these chronic stroke patients spontaneous improvement is not expected. Also measurements will be performed at 0 weeks and at 12 weeks post training. Primary outcome measures: Wolf Motor Function Test, ABILHAND, and Goal Attainment Scaling. Secondary outcome measures are: motricity index, plate tapping task, active range of motion, perceived strength and fatigue.

All tests and exercises are pain free, easy to perform, non-intrusive, and not high demanding. Subjects will have to undergo an assessment via a movement protocol as well as clinical outcome tests. In both cases the type of assessment does not differ from the normal assessment at a rehabilitation centre. The training is assisted using a robotic device called the Haptic Master (HM). The participant's arm is attached to the HM via a gimbal. The HM has a number of safety features to avoid overload on the participant's arm. Any potential risk for overload is minimised by a) careful build-up of the training regime; b) regular check-ups; and c) checking the participant's status prior to the next training session.

Inclusion Criteria:

• Clinically diagnosed with a central paresis of the arm/hand at entry in the study, caused by a supratentorial stroke;
• Post-stroke time more than 6 months (i.e. chronic phase after stroke);
• Having completed their active clinical rehabilitation program
• Hemiparesis featuring a low to moderate proximal (shoulder and arm) muscle strength:

Motricity Index shoulder/arm item: minimum score of 14 and maximum 25 (out of 33), corresponding to a maximum active shoulder abduction of up to 90 degrees without resistance; and/or a minimum active shoulder anteflexion of 30 degrees and a maximum active range of motion of 120 degrees shoulder joint anteflexion which can actively be maintained for 10 seconds;

• a fair cognitive level, i.e. being able to understand the questionnaires and measurement instructions;
• ability to read and understand Dutch.

Exclusion Criteria:

• Severe spasticity of the arm, i.e. Modified Ashworth Scale (MAS) score for the elbow and wrist flexors: each ≥ 3;
• Severe visual impairment and/or severe cognitive impairment which may interfere with the execution of the arm-hand tasks or the measurements;
• Severe neglect in the near extra personal space (38), established by the letter cancellation test (39) and Bell's test (quantitative evaluation, (40)) with a minimum omission score of 15% (Ferber, 2001);
• Broca aphasia, Wernicke aphasia, global aphasia: as determined by the Akense Afasie Test (AAT) (41);
• Severe apraxia as measured by the apraxia test of van Heugten (42);
• no informed consent.