This study aims to develop individualized therapeutic protocol to improve stroke recovery and fits the scheme of Mechanisms, biomarkers, and treatment strategies for stroke, aging, and Parkinson's disease by advanced EEG analysis and transcranial magnetic stimulation technique.

It has long been a challenge to promote the recovery after stroke. So far, rehabilitation has been the most evident way for this purpose, while pharmacological approaches have done very little.

However, the unsatisfied results urge us to recognize the need of new therapeutic solutions for stroke recovery. In the past few years, the development of non-invasive brain stimulation techniques, which are capable of altering the brain excitability through plasticity-like mechanisms for up to an hour or so, lead to the expectation of improving the recovery of stroke and other neurological disorders by manipulation neuroplasticity. Nevertheless, even with the help of brain stimulation, the results in stroke recovery are diverse and limited, although most studies agree that non-invasive brain stimulation, including repetitive transcranial magnetic stimulation (rTMS) is beneficial to the recovery after stroke.

Therefore, it is an important topic to ameliorate the recipe of brain stimulation for stroke recovery to improve its efficiency and efficacy. Facing this challenge, the investigators designed this study to develop individualized rTMS protocol for stroke recovery, to verify the predictor of stroke outcome discovered by DCM and to explore factors that cause the variability of rTMS. Based on our recent findings that revealed key patterns of stroke recovery base in EEG-based EEG based dynamic causal modelling (DCM), the individualize TMS protocol will be developed. The investigators will first apply different protocols of rTMS given to a variety of locations within the motor network to evaluate their effects on the DCM network. The results will be used for selecting proper rTMS protocols for stroke patients based on their DCM results. The effect of the individualized rTMS protocol will be tested in a small group of patient first. Then the longer therapeutic effect of the protocol will be assessed in a larger patient group under a double-blinded randomized control basis. With the help of this project, the investigators will be able to understand how different protocols of rTMS given to a variety of locations within the motor network affect the DCM network. Such information will be useful for developing individualized rTMS protocol for not only stroke therapy in the present study, but also other neurological disorders, e.g. Parkinson's disease and dystonia, in the further studies.

Furthermore, the investigators expect that the patients stimulated with rTMS using the protocol matched to their DCM network will recover better than those treated with unmatched protocol. This part of result could be further applied as a clinical application to facilitate stroke recovery. The predictors for stroke recovery revealed by the DCM model will also be verified in the current project. This will give us more information about the accuracy of the predictors and may give us new insights for the further development of stroke therapy. The investigators will further analyze EEG data and the results to provide knowledge of the state-dependent effect of brain on the variability of the current forms of rTMS and other noninvasive brain stimulation. This part of information will be very helpful for developing the strategy for eliminating the variability of rTMS and for improving the protocol of rTMS to be more powerful and efficient.

Inclusion Criteria:

• first-ever stroke
• in subacute stage (between 1-3 months after stroke onset)
• monohemispheric ischemic stroke
• able to follow the command of the shoulder flexion
• mild to moderate paresis in the upper limb (the motor arm score of National Institute of Health Stroke Scale (NIHSS) = 1 & 2 out of 4, where 4 = no movement)

Exclusion Criteria:

• being less than 20 years or over 75 years of age
• history of seizures or epileptiform discharges on electroencephalogram (EEG)
• pregnancy
• aphasia, apraxia, and concomitant neurological diseases or other severe medical diseases (e.g., sepsis, malignancy, hepatic or renal failure).
• having a pacemaker, an implanted medication pump, a metal plate in the skull, or metal objects inside the eye or skull (for example after brain surgery or a shrapnel wound)
• taking tricyclic antidepressants, neuroleptic agents and other drugs that lower the seizure threshold
• Insulin dependent diabetes
• Thyroid disease