Novel Brain Signal Feedback Paradigm to Enhance Motor Learning After Stroke

Stroke (795,000/year in the US and 30 million existing stroke survivors in the world) damages brain neural structures that control coordinated upper limb movement. To most effectively target the brain damage, interventions should be directed so as to restore brain control serving coordination of peripheral neuromuscular function. Currently, there is a lack of a transformative intervention strategy, and only limited efficacy is seen in response to neural rehabilitation that is only peripherally-directed (limbs e.g.) or only directed at the brain. This study will employ a novel neural feedback approach with a closed-loop, real-time paradigm to engage and retrain existing brain function after stroke. Real-time functional magnetic resonance imaging (rtfMIR) provides neural feedback with the advantage of precisely identifying the location of brain activity for multiple cognitive and emotional tasks. However, the rtfMRI is costly and precludes motor learning that requires sitting and engaging the upper limb in complex motor tasks during imaging acquisition. In contrast, real-time functional near-infrared spectroscopy (rtfNIRS), although not as spatially precise as rtfMRI, offers a low-cost, portable solution to provide brain neural feedback during motor learning. This proposal will utilize both technologies in a hybrid, sequential motor learning protocol. Moreover, the study protocol will also simultaneously involve both central effective signals (through neural feedback) and peripheral affective signals by employing neutrally-triggered functional electrical stimulation (FES)-assisted coordination practice, which produces peripherally-induced affective signals from muscle and joint receptors. This novel combination intervention protocol will engage the central nervous system, motor effective pathway training along with induction of affective signal production (FES-assisted practice), all of which will be implemented within the framework of evidence-based motor learning principles.

This study aims to develop and test an innovative protocol for recovery of wrist extension after stroke, using a combination of rtfMRI, rtfNIRS, FES, and motor learning. Aim I. Test the innovative coordination training protocol of combination rtfMRI/rtfNIRS central neural feedback and peripherally-directed, neurally-triggered FES-assisted coordination practice implemented within a framework of motor learning principles. Hypothesis 1. Chronic stroke survivors will show significant improvement in upper limb function in response to the combined rtfMRI/rtfNIRS central neural feedback; peripherally-directed FES-assisted coordination practice of wrist and finger extension; and whole arm/hand motor learning (Primary measure: Pre-/post-treatment change score in Arm Motor Abilities Test - function domain (AMAT - F); secondary measure: Pre/post-treatment change score in Fugl-Meyer upper limb coordination.

chronic stroke, functional MRI, functional NIRS, motor learning, FES, upper extremity, neural feedback, arm/hand functional training

Intervention: Stroke subjects will receive neural feedback plus FES and motor learning intervention that spans 3 phases and up to a total of 60 sessions. Phase I: real-time fMRI neural feedback training; Phase II: rtfNIRS-based neural feedback learning (built upon self-regulation strategies learned in Phase I and also assisted by neurally-triggered, peripherally-directed FES motor practice of wrist and finger extension); Phase III: motor learning minus neural feedback for an additional sessions up to 60 total; Phase IV: follow-up testing at 3 months after-treatment ends

We are not testing the feasibility of the imaging methods; that has been well established and is used clinically. We are testing the feasibility of using neural feedback clinical imaging methods in a neural feedback paradigm which involves sequential rtfMRI (phase I) and rtfNIRS (phase II) training; Neurally-triggered, peripherally-directed FES-assist practice of wrist and finger extension will be combined with rtfNIRS training in Phase II; up to 60 total sessions, including additional motor learning sessions without brain neural feedback will be provided in Phase III.

AMAT-F is a measure of 13 complex, coordinated tasks used in everyday living: functional normality of movement during the 13 tasks. Minimum clinically important difference (MCID) is 0.44 change score. AMAT-F : Arm Motor Abilities Test, functional domain. minimum = 0 points. maximum = 5 points. 5 points is normal function.

FM: Fugl-Meyer Coordination Scale: arm/hand coordination of isolated joint movement. The minimum clinically important difference (MCID) is 4.25 points. FM: 0 points, no movement; 66 points, normal coordination throughout the upper limb.

Inclusion Criteria: Cognition sufficiently intact to give valid informed consent to participate.* Sufficient endurance to participate in rehabilitation sessions. Ability to follow 2 stage commands. Medically Stable Age > 21 years. Impaired upper limb function as follows: impaired ability to flex and extend the wrist. At least 5 degrees of wrist flexion and extension of the wrist. Passive ROM of wrist extension of at least 20 degrees. At least 6 months post stroke. Exclusion Criteria: Metal implants, pacemaker, claustrophobia, inability to operate the MRI patient call button or any other contraindications for MRI. Acute or progressive cardiac (including cardiac arrhythmias), renal, respiratory, neurological disorders or malignancy. Active psychiatric diagnosis or psychological condition, or active drug/alcohol abuse. Lower motor neuron damage or radiculopathy. More than one stroke. Pregnancy (discontinued from the study, if a woman becomes pregnant). * The combined scores for the Aid to Capacity Evaluation (ACE) and Mini-Mental Status Examination (MMSE) as follows: MMSE 24-30 + the ACE score that states 'definitely capable' MMSE 17 - 23 + the ACE score that states 'probably capable'

De-identified, anonymized data-set will be created and shared after manuscript publication. Such data sets will be maintained locally according to the institutional policy and guidelines. De-identified data sets will be made available upon request to the study PI after study completion.

Matarasso AK, Rieke JD, White K, Yusufali MM, Daly JJ. Combined real-time fMRI and real time fNIRS brain computer interface (BCI): Training of volitional wrist extension after stroke, a case series pilot study. PLoS One. 2021 May 6;16(5):e0250431. doi: 10.1371/journal.pone.0250431. eCollection 2021.