This study integrates the wireless EEG system with an ordinary rehabilitation device (an upper limb ergometer, "arm bike") used in the Department of Physical Medicine and Rehabilitation at our hospital for a brain-computer-interface (BCI)-controlled neurorehabilitation device, and aims to test the effectiveness of this device. We hypothesize that, the coupling of electroencephalographic signals related with initiation of limb movements with a mechanical device which assists the intended movement is effective to facilitate motor recovery in patients with brain lesion. We propose to enroll 20 patients with cerebrovascular accident (CVA) (4-24 months after the onset of CVA) and the patients will be randomly assigned to experimental (using BCI controlled device and undergoing standard rehabilitation) and control groups (undergoing standard rehabilitation alone). To compare the rehabilitation results among these groups, we propose to use various assessment tools including clinical evaluation (Fugl-Meyer assessment, Modified Ashworth scale, Motor Activity Log, Functional Independence Measure) as well as functional Magnetic Resonance Imaging (fMRI) and Diffusion Tensor Imaging (DTI) before, immediate and 2 months after completion of the training protocol.

Evaluating the effectiveness of Wireless EEG-based BCI-controlled Rehabilitation System in patients with stroke

Applying the brain-computer interface (BCI) to improve the life-quality of handicaps and conveniences of healthy people in real life has been listed as one of the top 20 issues in the neuroscience field in recent 20 years. Over past years, the Biomedical Engineering R & D Center in China Medical University (CMU) and Hospital has devoted to develop wireless and wearable brain-signal detection equipment and the related software and hardware. Recently, the wireless electroencephalogram (EEG) system has been integrated and tested, side-by-side with a commercially available wired EEG system, which is oftentimes used as a standard in most laboratories for EEG experiments. After some examinations with cognitive tasks, the quality of the device and detected signals has been comparable to that of a commercial EEG system. As a result, we are further integrating the wireless EEG system with an ordinary rehabilitation device (an upper limb ergometer, "arm bike") used in the Department of Physical Medicine and Rehabilitation at our hospital for a BCI-controlled neurorehabilitation device, which we propose to use in the rehabilitation therapy for patients with stroke. We hypothesize that, the coupling of electroencephalographic signals related with initiation of limb movements with a mechanical device which assists the intended movement is effective to facilitate motor recovery in patients with brain lesion. To test the effectiveness of the proposed wireless EEG-based BCI-controlled rehabilitation device, we propose to enroll 20 patients with cerebrovascular accident (CVA) (4-24 months after stroke attach) and the patients will be randomly assigned to experimental and control groups. Patients in the experimental group will undergo 80 minutes of standard rehabilitation therapy and 20 minutes of BCI-controlled upper limb ergometer training during one rehabilitation session; those in the control group will take 100 minutes of standard rehabilitation therapy. All participants will receive 3 rehabilitation sessions each week for 8 weeks (a total of 24 sessions). To evaluate the rehabilitation result with different training protocols, we propose to use the behavioral assessment and brain imaging tools (fMRI and DTI). To compare the rehabilitation results among these groups, we propose to use various assessment tools including clinical evaluation (Fugl-Meyer assessment, Modified Ashworth scale, Motor Activity Log, Functional Independence Measure) as well as functional Magnetic Resonance Imaging and Diffusion Tensor Imaging before, immediate and 2 months after completion of the training protocol. If significant differences on behavioral and neuroimage evaluations between the two groups can be achieved, we will integrate the wireless-EEG rehabilitation system and behavioral-neuroimage assessment procedure as a new rehabilitation protocol for real clinical trial with a larger sample size.

Inclusion Criteria:

• cerebrovascular accident (CVA) for the first time
• between 4 months and 2 years after onset of CVA
• diagnosis of CVA proved by brain computed tomography or magnetic resonance imaging (MRI)
• motor status of CVA-affected proximal upper extremity: equals to or better than Brunnström stage IV
• ability to understand verbal commands and cooperate with test procedures

Exclusion Criteria:

• pain in the CVA-affected upper extremity, which adversely influences function
• severe joint contracture in the CVA-affected upper extremity
• strong spasticity (Modified Ashworth scale >3)
• poorly controlled epilepsy
• inability to undergo MRI for medical or other reasons