iBCI Optimization for Veterans With Paralysis

VA research has been advancing a high-performance brain-computer interface (BCI) to improve independence for Veterans and others living with tetraplegia or the inability to speak resulting from amyotrophic lateral sclerosis, spinal cord injury or stoke. In this project, the investigators enhance deep learning neural network decoders and multi-state gesture decoding for increased accuracy and reliability and deploy them on a battery-powered mobile BCI device for independent use of computers and touch-enabled mobile devices at home. The accuracy and usability of the mobile iBCI will be evaluated with participants already enrolled separately in the investigational clinical trial of the BrainGate neural interface.

After VA IRB approval, this VA RR&D study will engage participants in the BrainGate clinical trial (IDE, sponsor-investigator LR Hochberg). This study does not create a new clinical trial or modify the existing clinical trial as already listed on clinicaltrials.gov This project builds on a custom, mobile neural signal processing device with exceptional processing and low power characteristics, which has been developed through previous VA RR&D funded research. This project takes advantage of the exceptional processing system, previously developed and validated, to create and quantify advanced neural decoding algorithms that show promise (in preclinical studies) for improving the accuracy and reliability of neural decoding - but that are likely too computationally demanding to be viable on existing real-time BCI systems. Decoding methods will include magnitude kinematic decoding with recursive neural networks and high-dimensional discrete gesture decoding. Computational methods to be evaluated include latent space methods and stable manifolds to improve day-to-day reliability of high performance and high-dimensional orthogonalization approaches to improve the independence of kinematic and gesture decoding.

Spinal Cord Injury, Amyotrophic Lateral Sclerosis, Brain Stem Infarctions, Locked-in Syndrome, Muscular Dystrophy

Each participant enrolled in this early feasibility study is essentially an independent data point. The primary outcome of the IDE to which this VA study is attached is safety; feasibility is a secondary outcome. The outcome of this specific VA study, which engages participants in the BrainGate IDE, is usability of the mobile iBCI decoding system as measured through quantitative performance measures.

Performance of new decoding algorithms and methods will be developed and embedded in a small, mobile neural processor. The utility of these will be assessed separately with participants in the BrainGate pilot clinical trial, IDE.

An embedded neural signal processor device will be evaluated for its ability to provide accurate and reliable closed-loop control in a home-based brain-computer interface.

Rate of successful closed-loop acquisition of on-screen targets using imagined gestures to move a computer cursor or to select icons on a computer screen.

Inclusion Criteria: Inclusion criteria are extensive and are determined by the associated BrainGate IDE(clinicaltrials.gov # NCT00912041) Informally, participants will be tetraplegic or anarthric with little or no functional use of the arms and legs Exclusion Criteria: Exclusion criteria are extensive and are determined by the associated BrainGate IDE(clinicaltrials.gov # NCT00912041).

Certain clinical data collected in this study are subject to institutional policies and local, state, and Federal laws and regulations, including the HIPAA Privacy Rule, which limit the distribution of such data in order to maintain the rights and privacy of research participants. Figuring out how to truly de-identify datasets, particularly given the range of potential neurological and psychiatric insights that one could, in the future, make from these data, is a considerable ethical challenge. Limited data resulting from this study will be released in association with peer-reviewed manuscripts to provide for independent validation of manuscript results and methods by others in the field. Due to the unique nature of this human trial data, it is expected that consultation with the authors will be required for external parties to perform accurate, appropriate, and meaningful secondary analyses.

This VA study does not create or modify a clinical trial. This Links to the associated clinical trial (IDE) which this particular VA study leverages.