A brain-computer interface (BCI) is a system that establishes a direct communication channel between neural activity in the brain and an external device — bypassing the normal pathway through nerves and muscles. BCIs read electrical signals generated by neurons (or the collective activity of neural populations), decode them into commands, and use those commands to control computers, robotic limbs, communication devices, or stimulation systems.

Core Components

Every BCI system has three fundamental components:

  1. Signal acquisition: Electrodes or sensors that capture neural activity, ranging from non-invasive scalp EEG to fully implanted intracortical arrays.
  2. Signal processing and decoding: Algorithms that translate raw neural signals into meaningful commands — for example, inferring intended cursor direction from motor cortex spike patterns.
  3. Output device: The controlled interface, which may be a computer cursor, a speech synthesizer, a robotic arm, or a functional electrical stimulation system.

Invasive vs. Non-Invasive

BCIs exist on a spectrum of invasiveness:

  • Non-invasive: EEG headsets placed on the scalp. High patient accessibility, no surgical risk, but limited spatial resolution and signal quality due to signal attenuation through skull and scalp tissue.
  • Minimally invasive: Endovascular approaches (e.g., Synchron's Stentrode) placed via blood vessel catheterization; no open brain surgery required.
  • Invasive (ECoG): Electrode arrays placed on the cortical surface via craniotomy. Better signal quality than EEG without penetrating brain tissue.
  • Fully invasive (intracortical): Electrodes inserted directly into brain tissue (e.g., Neuralink N1, Utah Array). Highest signal quality and spatial resolution but requires neurosurgery and carries highest procedural risk.

Clinical Applications

Current FDA-regulated clinical BCI applications include:

  • Motor restoration and communication: Neuralink PRIME study, Synchron SWITCH/COMMAND trials — enabling individuals with ALS or paralysis to control computers by thought.
  • Closed-loop therapeutic stimulation: NeuroPace RNS System for epilepsy; Medtronic Percept PC and adaptive DBS for Parkinson's disease.
  • Rehabilitation: BCI-controlled exoskeletons (IpsiHand) for post-stroke motor recovery.

History

The term BCI was coined by Jacques Vidal at UCLA in 1973, who published the first formal description of a system for real-time computer control using EEG signals. The first human intracortical BCI was implanted in Matthew Nagle in 2004 by the BrainGate consortium at Brown University, enabling cursor control using motor cortex recordings from a Utah Array. The field has advanced dramatically since, with Neuralink's N1 implant representing the current state-of-the-art in high-channel-count wireless intracortical BCI as of 2024.