In rehabilitating chronic motor-impaired stroke survivors with a brain computer interface (BCI), there is currently a fundamental limitation in tailoring the neuroprosthetic strategy to the patient specific pathology. The current barrier for further advancement of BCI-driven neurorehabilitation is that the mechanisms behind BCI-based approaches to motor recovery are not well understood, thus hindering the design of effective and individualized rehabilitative methods [1-4]. The long-term goal of this research is to reliably restore function to patients with a chronic stroke-induced motor deficit by aligning their BCI rehabilitative strategy with their specific stroke- induced pathophysiology. Thus, the mechanism in which BCIs alter the brain is of substantial scientific and clinical importance. The objective of this proposal is to define the changes in the brain of chronic subcortical hemiparetic stroke patients that occur with BCI-driven rehabilitation that uses contralesional motor signals. Our central hypothesis is that ongoing use of a BCI-controlled robotic hand exoskeleton, driven by contralesional motor signals produced during motor imagery of the affected hand, will induce functional remodeling that will correlate with motor recovery. The rationale for this work is that the knowledge will create a mechanism-driven approach to neuroprosthetic solutions for stroke in the future. Guided by strong evidence that a BCI using the unaffected hemisphere can improve motor function in chronic stroke patients [5], we will test the central hypothesis with the following three specific aims: 1) Define corticospinal tract (CST) remodeling associated with the functional improvement induced by BCI-driven rehabilitation using motor signals from the unaffected hemisphere in chronic hemiparetic stroke patients., 2) Delineate the remapping of task-based activations that occurs with contralesionally-driven, BCI-mediated functional recovery in chronic hemiparetic stroke patients, and 3) Define the alteration in motor network connectivity associated with the functional improvement that is induced by contralesionally-driven, BCI-rehabilitation in chronic hemiparetic stroke patients. Under the first aim, we will use diffusion tensor imaging (DTI) to evaluate the alterations in fractional anisotropy (FA) that occur within the CST before and after BCI therapy using contralesional motor signals. In the second aim, we will evaluate the changes in topographical cortical activations that occur through BCI therapy. Under the third aim, we will evaluate the alterations in hemispheric network interactions by using fMRI to define changes in resting state functional connectivity. This project is innovative because it is a substantial departure from the status quo by expanding the role the unaffected hemisphere and bihemispheric interactions can play in BCI-mediated rehabilitation. The proposed research will be significant because the knowledge will vastly improve the characterization of how a BCI can alter both anatomy and function in a chronic stroke patient's brain and subsequently how these changes can be targeted for a tailored neuroprosthetic intervention. Ultimately, this will inform the development of novel treatments for stroke patients.
The proposed research is relevant to public health because developing a deeper understanding of how the brain changes with neuroprosthetic rehabilitation of chronic stroke will fundamentally expand the knowledge of how to best harness these dynamics with a brain computer interface to induce a functional recovery. Thus, the proposed research is relevant to the NINDS mission, which is to develop knowledge about the brain and to use that knowledge to reduce the burden of the neurologic disease of stroke.
