Chronic Stroke Rehabilitation With Contralesional Brain-Computer Interface

The Neural Mechanisms of a Contralesionally-Driven Brain-Computer Interface for Motor Rehabilitation of Chronic Stroke

The purpose of this research study is to show that a computer can analyze brain waves and that those brain waves can be used to control an external device. This study will also show whether passive movement of the affected hand as a result of brain-based control can cause rehabilitation from the effects of a stroke. Additionally, this study will show how rehabilitation with a brain-controlled device may affect the function and organization of the brain. Stroke is the most common neurological disorder in the US with 795,000 strokes per year (Lloyd-Jones et al. 2009). Of survivors, 15-30% are permanently disabled and 20% require institutional care (Mackay et al. 2004; Lloyd-Jones et al. 2009). In survivors over age 65, 50% had hemiparesis, 30% were unable to walk without assistance, and 26% received institutional care six months post stroke (Lloyd-Jones et al. 2009). These deficits are significant, as recovery is completed after three months (Duncan et al. 1992; Jorgensen et al. 1995). This large patient population with decreased quality of life fuels the need to develop novel methods for improving functional rehabilitation. We propose that signals from the unaffected hemisphere can be used to develop a novel Brain-Computer interface (BCI) system that can facilitate functional improvement or recovery. This can be accomplished by using signals recorded from the brain as a control signal for a robotic hand orthotic to improve motor function, or by strengthening functional pathways through neural plasticity. Neural activity from the unaffected hemisphere to the affected hemiparetic limb would provide a BCI control in stroke survivors lesions that prevent perilesional mechanisms of motor recovery. The development of BCI systems for functional recovery in the affected limb in stroke survivors will be significant because they will provide a path for improving quality of life for chronic stroke survivors who would otherwise have permanent loss of function. Initially, the study will serve to determine the feasibility of using EEG signals from the non-lesioned hemisphere to control a robotic hand orthotic. The study will then determine if a brain-computer interface system can be used to impact rehabilitation, and how it may impact brain function. The system consists of a research approved EEG headset, the robotic hand orthotic, and a commercial tablet. The orthotic will be made, configured, and maintained by Neurolutions. Each participant will complete as many training sessions as the participant requires, during which a visual cue will be shown to the participant to vividly imagine moving their impaired upper extremity to control the opening and closing of the orthotic. Participants may also be asked to complete brain scans using magnetic resonance imaging (MRI).

The purpose of this research study is to show that a computer can analyze brain waves and that those brain waves can be used to control an external device. Additionally, this study will show whether passive movement of the affected hand as a result of brain-based control can cause rehabilitation from the effects of a stroke. Stroke is the most common neurological disorder in the U.S. with 795,000 strokes per year (Lloyd-Jones et al. 2009). Of survivors, 77% experience weakness of loss of motor function in the upper limb (Lawrence et al, 2001). Motor recovery in post-acute stroke patients is complicated by an apparent plateau in the ability to achieve recovery beyond 3 months after stroke (Duncan et al. 1992; Jorgensen et al. 1995; Lloyd-Jones et al. 2009). The large patient population with decreased quality of life and requiring significant medical resource fuels the urgent need to develop novel methods for improving functional rehabilitation in chronic stroke survivors. We propose that cortical signals from the unaffected hemisphere of chronic stroke survivors can be used to control Brain-Computer Interface (BCI) system to facilitate functional recovery. The development of such rehabilitative BCI systems is significant because it provides a path to functional recovery currently unavailable to many chronic stroke patients. Subject Selection: Participants will be recruited from patient populations of collaborators and colleagues of the principal investigator, as well as from previous research studies of the principal investigator and colleagues. Participants from previous research studies will also be recruited. Colleagues will provide study information to interested candidates, and candidates will contact the research team if they will to be screened by study staff. Patients will be asked a series of screening questions to determine their eligibility for the study. Study Methods Study staff will screen participants for eligibility after completing the informed consent process and documentation. After screening for study eligibility, the participant will be assigned to group 1 or group 2. Group 1 participants will complete the study with the addition of functional MRI imaging (fMRI) and a range of motion (ROM) home exercise program appropriate for the patient as determined by a clinical specialist. Group 1 participants will undergo 12 weeks each of device use and the home exercise program. Half of the participants will use the device before the ROM program, and half afterwards (i.e. a crossover design). Neuroimaging scans will take place immediately prior to device use, immediately prior to beginning ROM therapy, and at study completion. Group 2 participants participate in the study with no imaging. Following eligibility screening, participants will be placed into Group 1 or 2 based on ability and willingness to participate in MRI scans. The study will be run in two phases. In Phase 1, each participant will complete up to 3 sessions for recording EEG signals. For recording, patients will wear a research grade EEG headset with EEG electrodes in place. The signals will be recorded with brain computer interface software. If the participant's EEG signals are adequate for controlling a BCI mediated hand orthosis, they will continue onto phase 2 of this study. Phase 2 of this study, participants will be issued a BCI mediated hand orthotic and be advised to use the device daily at home (5 out of 7 consecutive days) for a minimum of 12 weeks. The participant will have motor assessments at 4 weeks, 8 weeks and 12 weeks of device use. Group 1 patients follow the same schedule for ROM therapy and continued motor function assessments. Should the participant be achieving progress as evidenced by motor assessments at 12 weeks of device use, the participant may be asked to continue to utilize this therapy daily at home until progress plateau's. Should this occur, participants would have motor assessments completed every 4 weeks of additional device use (beyond the minimum 12 weeks). Participants may be contacted to complete a motor screening 6 months post device use to assess durability of motor recovery of the affected upper extremity. Phase 1 - EEG Signal Assessment Participants will complete two EEG screening visits to find a consistent EEG signal to control the robotic hand orthosis. Participants may be asked to complete a third EEG screening if a consistent signal could not be identified in the first two EEG screenings. During EEG screenings, participants will wear an EEG headset with a subset of the standard 10-20 system of electrode coordinates. Proper connection will be verified using signal inspection. A 7.5 minute set of resting data will be recorded. Secondly, a motor imagery task will be completed in which participants will receive visual cues/prompts on a computer screen to imagine finger tapping movements of the left hand, right hand, or both hands for a period of 5-10 seconds per cue. Screening data will be analyzed to ensure that sufficient cortical signals are present for device control. Spectral power changes in the unaffected hemisphere will be analyzed using an r-squared analysis. Participants must exhibit significant power changes at electrode locations of the motor and pre-motor cortical areas to achieve device control. MRI Methods: We will obtain structural, diffusion tensor imaging (DTI) and resting state fMRI (rs-fMRI) from the patients in this study. Structural images are acquired using T1- and T2-weighted scans. DTI scans will be acquired with b-values of 300, 1000 and 2000 s/mm² and 8, 32, and 60 diffusion directions. Tract-specific measurements, such as volume, radial diffusivity, and fractional anisotropy (FA) will be obtained from the DTI data. We will collect approximately 22.5 minutes of rs-fMRI (3 scans at 7.5 minutes each). The data will undergo a standard preprocessing stream including spatial smoothing, temporal band pass filtering and removal by regression of sources of spurious variance. Data will be volume-censored to avoid motion-induced artifacts. Phase 2 - BCI Therapy & Motor Assessments After completing EEG screenings, participants in Group 1 and Group 2 will complete two sets of motor assessments in-office on two separate days to establish a baseline. These assessments to be completed are: UEFM Motricity Index Modified Ashworth Scale (Elbow Flexion, Wrist Flexion) Gross Grasp Hand Dynamometer Arm Motor Ability Test (AMAT) Participants will be given a device and trained in its use and care by a clinical specialist following the initial motor assessments. At each follow-up motor assessment visit, the clinical specialist will have the participant bring their device in each follow up. The clinical specialist will download all data stored on the device, address any questions or concerns, modify the fit of the, and modify the home exercise program as needed. During the final motor assessment visit, participants will be asked to complete a final EEG screen identical to the baseline EEG screen in order to study robust changes in cortical activity that may have occurred during BCI rehabilitation. The participant will also be asked to complete a patient experience survey as it relates to user experience of the device. Patients will be reimbursed for their participation in the study. EEG screenings and motor evaluations will be reimbursed at a rate of $25.00 each. MRI scans will be reimbursed at a rate of $50.00 each. Patients who complete the study on protocol will receive a bonus $50.00 incentive. Risk / Safety Information (Device Use): Likely: None Less Likely: Fatigue from repetitive computer tasks and/or frustration. Eyestrain and fatigue could result from prolonged attempts, watching the computer screen throughout. Rare: Minor discomfort associated with muscle stimulation. A research staff member will be available while the subject is participating in any portion of the research study. During experimentation, patient will be reminded that they may stop the study at any time and they can delay or terminate any sessions if they are experiencing discomfort. The battery power of the EEG headset is very low and thus presents no risk. The robotic hand orthosis will be operated to move the participants hand within their physiologic range of motion and will not exert forces great enough to physically harm the patient. Additionally, participation is voluntary, and the individual may choose to terminate at any time. Risk / Safety Information (MRI): Likely: Mild - Fatigue, Discomfort from lying in the scanner. Less Likely: Mild - Feeling of claustrophobia. Discomfort from loud noise of the scanner. Rare: Life Threatening - Injury from metal object in the body. There is substantial risk to persons who have metallic objects inside their bodies, since the MRI scanner uses a high strength magnet. Examples of these include surgical staples left in the body following surgery, middle ear prostheses or cochlear implants, permanent eye liner, metal foreign objects lodged inside the eye, heart pacemakers, and pins inside the knees or other joints. If patients have any kind of metallic object in their body not been tested for MRI safety, they may not participate in the MRI portion of the study. Study Oversight: The decision to participate in this study is voluntary. The participant may choose to not participate or may withdraw from the study for any reason without penalty or loss of benefits to which are otherwise entitles and without any effect on future medical care. The principal investigator or the sponsor can stop one's participation at any time without the participant's consent for any reason. Some reasons may include, but are not limited to: If it appears to be medically harmful to the participant; If the participant fails to follow directions for participating in the study; If it is discovered that the participant does not meet the study requirements; If the study is canceled; or For administrative reasons, including competitive enrollment - the target number of subjects has entered the study The study doctor / principal investigator of the study will provide oversight throughout the clinical trial. Data Analysis / Management: The data will be analyzed with multiple techniques, including: Multiple mathematical algorithms will be used to translate raw analog electrocortical activity into a statistically significant signal profile. These would include such approaches as the autoregressive analysis, fast fourier transforms, analysis of variance, signal-to-noise ratio (SNR) technique, Cross Correlation Signal Technique, Etc. These methods would primarily be used for offline analysis of signals. In phase 2, tests of motor function (UEFM, Motricity Index, Modified Ashworth, grasp strength, during use of the device, and post using the device will be used to test the impact of the device on the rehabilitative effects of upper extremity function. In phase 2, changes in functional connectivity patterns before and after treatment will be assessed using statistical analysis as follows: construction of regions of interest (ROI); ROI will be used as "seeds" to create ROI to whole brain voxel-wise correlation maps; correlation map statistical methodologies (analyses that treat participants as a random effect to test for group effects in the Fischer-z transformed correlation maps, and resulting group maps corrected for multiple comparisons using previously computed Monte-Carlo simulations). The UEFM will serve as the primary measure of statistical success. Because we will use baseline tests before treatment begins, we will use a repeated measures test, such as paired t-test assuming a normal distribution of scores. The Upper Extremity portion of the UEFM assess grasp motor function in the affected upper extremity. The key statistical outcome will be based on the subtest grasp as the device has the greatest potential impact on this domain. Power estimates suggest that 10 subjects will be sufficient for significant results in a paired T-test; up to 20 patients will be enrolled to allow for attrition. Confidentiality: All clinical and experimental data will be de-identified by being assigned a randomly generated code. Additionally, all traceable data from copied medical records will also be removed. Paper records will be kept in a locked cabinet in a locked office suite. Electronic records will be stored on a lab computer in a password-protected file. Only study team members will have access to records. After the research project is completed, the principal investigator will delete all electronic files and shred any paper forms containing identifiers. A member of the research team may discuss the study with the participant in person or by phone to describe the study to the patient and determine if they are willing to participate. However, the patient will be consented in person. Project Goal: The ultimate goal of this project is to develop a functioning and clinically feasible method for restoring function to motor impaired stroke survivors. In developing a new rehabilitation method, we hope to create a system that allows for closed loop feedback through a robotic hand orthosis on the motor impaired side of stroke patients in response to intended movements of the muscles. The method, if successful would represent a non-invasive method of promoting motor learning and recovery in stroke survivors.

Study Population in 2 groups: Group 1 participates in MRI before treatment, at crossover, and at study completion. Group 1 participants either receive rehabilitation via BCI device then cross over to a standard range-of-motion program, or start with a range-of-motion program then crossover to receive BCI rehabilitation. A balanced number of participants will be assigned to the different orders within Group 1. Group 2 receives no MRI, and is not assigned a range-of-motion program. Thus, Group 2 only receives BCI rehabilitation and does not cross over.

Patients trained on use of BCI-controlled orthotic device are given a device for home use. Patients are asked to use the device an hour per day, 5 days per week, for 12 weeks. During device use, patients are instructed via pre-programmed instructions on a tablet paired with the device to either rest or vividly imagine moving their affected hand. The device receives signals from a scalp electrodes within a headset the patient dons prior to use. The device interprets these signals and closes the patient's hand during a successful rest trial, and opens the patient's hand during a successful move trial.

Active and Passive Range-of-Motion (AROM, PROM) therapy strategies are commonly prescribed by physical therapists for at-home post-stroke motor deficit rehabilitation that can be performed independently. Patients practice movement with joints and limbs affected by the stroke, either by using the unaffected limb (or the assistance of a caretaker) to stretch the affected limb (PROM) or by actively moving the affected limb (AROM). Patients are asked to perform this therapy one hour per day, 5 days per week, for 12 weeks.

Patients use electroencephalography (EEG) signals to control a motorized glove worn on their affected hand. The glove moves the patient's hand according to the type of signal detected (Rest vs Motor Imagery).

Patients repeatedly move or stretch the joints and muscles of their affected limb, either by actively moving the limb or assisting the limb with no active motion.

The primary outcome for determining motor function improvement is the change over time in the upper extremity portion of the Fugl-Meyer Assessment (FMA). The difference between FMA scores pre- and post-BCI rehab, subtracted by the change in FMA during range-of-motion therapy, will be used to quantify change in motor function.

Difference in fractional anisotropy (FA) of the corticospinal tract during BCI rehabilitation subtracted by change in FA of the corticospinal tract during range-of-motion therapy

Change in average resting state connectivity between left and right somatomotor brain regions during BCI rehabilitation subtracted by change in average resting state connectivity between left and right somatomotor brain regions during range-of-motion therapy

Change in spasticity measured with Motricity Index during BCI rehabilitation subtracted by change in spasticity measured with Motricity Index during range-of-motion therapy

Change in grasp strength during BCI rehabilitation subtracted by change in grasp strength during range-of-motion therapy

Change in ability to perform activities of daily living (ADLs) as measured by the Arm Motor Ability Test (AMAT) score during BCI rehabilitation subtracted by change in ability to perform ADLs as measured by AMAT score during range-of-motion therapy

Inclusion Criteria: Chronic stroke survivors at least 6 months post-stroke with moderate functional impairment of the right or left upper extremity as evidenced by motor function screening assessments If receiving Botox injections in the upper extremity for spasticity management, device use must be initiated within 15 days of a Botox injection Exclusion Criteria: Cognitive impairment as indicated by a Short-Blessed Test score of 8 or more Joint contractures in the affected wrist or digits Receptive aphasia or inability to follow written instructions as indicated by a score of 6 or less on the Mississippi Aphasia Screening Test High spasticity as indicated by a Modified Ashworth Scale of elbow flexion of 3 or greater Unilateral visual inattention (i.e. "neglect") as determined by unilaterally omitting 3 or more targets on the Mesulam Cancellation Test Patients contraindicated for MRI imaging due to safety concerns will be excluded from Group 1, but will have the option to be assigned to Group 2 should they meet other Inclusion and Exclusion criteria. Inability to produce EEG signals sufficient for device control following EEG screening

Anonymized participant demographics, EEG data, and neuroimaging data will be provided with other researchers by request starting 6 months after publication of primary findings.

Data becomes available 6 months after publication of primary findings. Data will be available indefinitely.

Data will be provided via secure transfer service upon request members of respected research institutions, be they academic, government, or otherwise.

Duncan PW, Goldstein LB, Matchar D, Divine GW, Feussner J. Measurement of motor recovery after stroke. Outcome assessment and sample size requirements. Stroke. 1992 Aug;23(8):1084-9. doi: 10.1161/01.str.23.8.1084.

Lloyd-Jones D, Adams R, Carnethon M, De Simone G, Ferguson TB, Flegal K, Ford E, Furie K, Go A, Greenlund K, Haase N, Hailpern S, Ho M, Howard V, Kissela B, Kittner S, Lackland D, Lisabeth L, Marelli A, McDermott M, Meigs J, Mozaffarian D, Nichol G, O'Donnell C, Roger V, Rosamond W, Sacco R, Sorlie P, Stafford R, Steinberger J, Thom T, Wasserthiel-Smoller S, Wong N, Wylie-Rosett J, Hong Y; American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics--2009 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation. 2009 Jan 27;119(3):e21-181. doi: 10.1161/CIRCULATIONAHA.108.191261. Epub 2008 Dec 15. No abstract available. Erratum In: Circulation. 2009 Jan 27;119(3):e182. Circulation. 2010 Jul 6;122(1):e11. Circulation. 2011 Oct 18;124(16):e424.

Lawrence ES, Coshall C, Dundas R, Stewart J, Rudd AG, Howard R, Wolfe CD. Estimates of the prevalence of acute stroke impairments and disability in a multiethnic population. Stroke. 2001 Jun;32(6):1279-84. doi: 10.1161/01.str.32.6.1279.

Jorgensen HS, Nakayama H, Raaschou HO, Vive-Larsen J, Stoier M, Olsen TS. Outcome and time course of recovery in stroke. Part II: Time course of recovery. The Copenhagen Stroke Study. Arch Phys Med Rehabil. 1995 May;76(5):406-12. doi: 10.1016/s0003-9993(95)80568-0.