This project aims at gaining insight into the role of sleep in motor learning and the first to apply sleep related learning methods in a rehabilitative setting. The primary objective is to investigate whether there is an effect of acoustic cueing during slow wave sleep on a specific motor task. The second objective is to investigate whether this effect can also be transferred to generalized motor rehabilitation outcome. Patients with an impairment of the arm function are randomized to either receiving the intervention ("cueing group") or to the control group. The patients are all listening to a melody during motor training but only the patients of the cueing group are also exposed to the melody during subsequent slow wave sleep. Performance gain is measured using an instrumented arm orthosis which is used for functional arm therapy over the sudy period of 5 consecutive days as well as during standard clinical assessments.

This is an open prospective pilot study. We plan to include 12 patients with stroke and to assign them into two groups (chosen by lot).

The patients will be tested to investigate the effect of cueing during slow-wave sleep on motor skill learning as well as on the general motor rehabilitation outcome (5 days). The study will be conducted in a controlled design while 6 patients will receive cueing whereas the other 6 control patients receive no cueing.

Study design: Effect of motor task-related acoustic cueing and motor recovery over 5 days of training On day 1 patient will have to perform different standardized short training tasks using the ArmeoSpring® system. Each of these tasks is designed to train speed and accuracy of movements whilst a leitmotif melody is played during the performance of the respective task. Every task will be performed twice at an easy and twice at a more difficult level. If one patient succeeds to reach a defined height of score at level 2 the patient will start with level 2 the next day and proceed to level 3 in order not to lose any measurable effect. The melody lasts for 5 to 10 seconds (similar to the design chosen by Antony et al 2012). The patients' movements will be measured during day and night with movement sensors, and during the training and clinical assessment sessions. Information about arm movement in relation to trunk movement is sent to a PC by wire transmission. Arm movements are translated into movements of an avatar (e.g. an arm-like structure) on the computer screen. Sensor measures will be transmitted via wire connection to a receiver, which is connected to a standard personal computer. Thus, this kinematic motion analysis system reconstructs all movements of the tested arm in real time. By moving this structure, the test subject can therefore fulfil a short task in a videogame-like environment (e.g. using technical items presented on the screen). Scores are given and visually presented to the patient.

During the following night patients will undergo electroencephalography (EEG) for the first time. Immediately before scheduled regular sleep hours, the EEG-Electrodes and the headphones (sleepphones® Classic) will be installed and EEG system (Embla® Titanium) will be started. The patients will sleep in their own normal hospital bed at the rehabilitation ward.

After onset of slow-wave-sleep as confirmed by online EEG, the melody will be played to the patients in the cueing group via headphones (sleepphones® Classic) in faint intensity (in-sleep-cueing without waking up the patient). Conversely patients of group 2 won't be exposed to the melody during sleep. Sound intensity will be similar to that of the background noise (approximately 35 dB sound pressure level). Once slow wave sleep finishes, the stimulation will also be stopped. Under ideal circumstances, the stimulation will be repeated in each slow-wave sleep phase. On day 2 - 5 the same motor tasks will be performed at the ArmeoSpring® device, with exposure to the respective melodies. The improvement in Scores will be will be measured.

During the following three nights the abovementioned procedures will be repeated.

On day 5 the motor tasks will again be performed at the ArmeoSpring® device without exposure to the respective melodies. The task scores will again be recorded and also the scores from the motor assessments will be taken in order to assess the general motor improvement over five days.

Inclusion Criteria:

• Male and Female patients above 18 years of age
• signed informed consent after being informed
• subacute (i.e. symptom onset 3 - 90 days ago) lateralized ischemia or hemorrhage as confirmed by brain imaging (CT or MRI).
• Fugl-Meyer Upper Extremity Score between 15 and 59

Exclusion Criteria:

• other neurological or systemic disorder which is likely to cause dementia, cognitive dysfunction or central motor symptoms
• severe sensory aphasia
• preexisting arm paresis
• intake of sedatives, or neuroleptics
• Relevant hearing loss
• Vulnerable Person