Demonstration of a portable intracortical brain-computer interface

Jeffrey M. Weiss, Robert A. Gaunt, R. Franklin, Michael L. Boninger, Jennifer L. Collinger

Abstract

While recent advances in intracortical brain-computer interfaces (iBCI) have demonstrated the ability to restore motor and communication functions, such demonstrations have generally been confined to controlled experimental settings and have required bulky laboratory hardware. Here, we developed and evaluated a self-contained portable iBCI that enabled the user to interact with various computer programs. The iBCI, which weighs 1.5 kg, consists of digital headstages, a small signal processing hub, and a tablet PC. A human participant tested the portable iBCI in laboratory and home settings under an FDA Investigational Device Exemption (NCT01894802). The participant successfully completed 96% of trials in a 2D cursor center-out task with the portable iBCI, a rate indistinguishable from that achieved with the standard laboratory iBCI. The participant also completed a variety of free-form tasks, including drawing, gaming, and typing.

The Challenge of Lab-Bound BCIs

Intracortical BCIs (iBCIs) record neural signals directly from the motor cortex to restore function for individuals with paralysis. Despite their potential, these systems have historically been "lab-bound," requiring massive racks of signal processors, multiple high-performance PCs, and complex analog cabling. This hardware footprint represents a significant barrier to clinical translation. For iBCIs to become practical assistive devices, they must transition from the laboratory to the user's home environment without sacrificing performance.

A Compact Hardware Solution

To address this, researchers developed a self-contained, 1.5 kg portable iBCI. The system replaces the traditional laboratory stack with three streamlined components:

  1. Digital Headstages: Two Blackrock Cereplex-E headstages digitize neural signals at the pedestal, significantly reducing environmental noise.
  2. wNSP Hub: A wearable signal processing hub that aggregates data from up to 256 electrodes.
  3. Tablet PC: A medical-grade Getac RX10H tablet running Windows 10, which serves as the central processing unit.
Comparison of the portable iBCI system (a) with the traditional laboratory hardware stack (b).

By digitizing signals at the source, the team achieved a 20% reduction in baseline noise and a 5% improvement in the signal-to-noise ratio (SNR) compared to traditional analog systems.

Software Architecture

The software was transitioned from a distributed multi-PC environment to a modular suite running entirely on the tablet. Utilizing the Dragonfly messaging architecture, the system coordinates real-time spike processing, decoding via an optimal linear estimator, and task execution.

Modular software architecture of the portable iBCI system.

This consolidation reduced processing delays by eliminating network transmission between PCs, though it introduced some timing jitter due to the non-real-time nature of the Windows 10 operating system.

Performance and Versatility

The system was evaluated with a 28-year-old participant with a C5/C6 spinal cord injury. Performance was remarkably stable: the participant achieved a 95.0% success rate at home and 97.5% in the lab using the portable system, compared to 95.8% on standard lab equipment.

Beyond controlled cursor tasks, the participant successfully engaged in several "free-form" activities:

  • Communication: Typing messages on an on-screen keyboard.
  • Entertainment: Playing computer games against able-bodied opponents using a gamepad emulator.
  • Creativity: Creating digital art in TuxPaint.

Conclusion and Future Directions

This study demonstrates that high-performance iBCI control can be maintained in a portable, home-ready format. While the system currently requires some supervision for setup, it marks a major step toward independent use. Future work will focus on developing user-centric interfaces for caregivers and upgrading hardware to support higher-bandwidth data for next-generation electrode arrays.

Summary updated

Abstract

Abstract While recent advances in intracortical brain-computer interfaces (iBCI) have demonstrated the ability to restore motor and communication functions, such demonstrations have generally been confined to controlled experimental settings and have required bulky laboratory hardware. Here, we developed and evaluated a self-contained portable iBCI that enabled the user to interact with various computer programs. The iBCI, which weighs 1.5 kg, consists of digital headstages, a small signal processing hub, and a tablet PC. A human participant tested the portable iBCI in laboratory and home settings under an FDA Investigational Device Exemption ( NCT01894802 ). The participant successfully completed 96% of trials in a 2D cursor center-out task with the portable iBCI, a rate indistinguishable from that achieved with the standard laboratory iBCI. The participant also completed a variety of free-form tasks, including drawing, gaming, and typing.
BibTeX
@inproceedings{weiss2019demonstration,
  author = {Jeffrey M. Weiss and Robert A. Gaunt and R. Franklin and Michael L. Boninger and Jennifer L. Collinger},
  title = {{Demonstration of a portable intracortical brain-computer interface}},
  booktitle = {Brain-Computer},
  year = {2019},
  doi = {10.1080/2326263x.2019.1709260},
}
Figure