Associative brain computer interface training increases wrist extensor corticospinal excitability in subacute stroke patients

In a recently developed associative rehabilitative brain computer interface system, electroencephalography is used to identify the most active phase of the motor cortex during attempted movement and deliver precisely timed peripheral stimulation during training. This approach has been demonstrated to facilitate corticospinal excitability and functional recovery in patients with lower limb weakness following stroke. The current study expands those findings by investigating changes in corticospinal excitability following the associative BCI intervention in post-stroke patients with upper limb weakness. In a randomized controlled trial, 24 subacute stroke patients, subdivided into an intervention group and a "sham" control group, performed 30 wrist extensions. The intervention comprised 30 pairings of single peripheral nerve stimulation at the motor threshold, timed so that the generated afferent volley arrived at the motor cortex during the peak negativity of the MRCP, which was identified using EEG. The sham group underwent the same intervention, though the intensity of the nerve stimulation was below the perception threshold. Immediately after training, patients in the associative group exhibited significantly larger amplitudes of muscular evoked potentials, compared to pre-training measurements in response to transcranial magnetic stimulation. These changes persisted for at least 30 minutes, and were not observed in the sham group. We demonstrate that motor evoked potential amplitudes increased significantly following paired associative BCI training targeting upper limb muscles in subacute stroke patients, which is in line with results from lower limb muscles.