Brain-computer interface (BCI)

Also: BCI, brain-machine interface, BMI, neural interface

By neuraspeak editorial · Updated 2026-10-07 · 1 min read

A brain-computer interface (BCI) is a system that records activity from the brain, translates it with software into commands, and uses those commands to control an external device such as a computer cursor, robotic arm, or speech synthesizer, without relying on the body's muscles or nerves.

How does a brain-computer interface work?

Every BCI has the same basic chain. Sensors pick up electrical (or sometimes blood-flow) signals from the brain. A computer cleans and summarizes those signals into features, such as how often neurons fire or how much power appears in a frequency band. A decoder, usually a machine-learning model trained on the user's own data, maps those features to an intended output: a cursor movement, a letter, a word, or a sound. The user sees or hears the result, which lets them adjust in real time.

The sensors are the main dividing line between BCI types. Noninvasive systems use electroencephalography (EEG) electrodes on the scalp. Invasive systems place electrodes on the brain's surface (electrocorticography, or ECoG), inside a blood vessel next to the brain (endovascular), or into the cortex itself (intracortical microelectrode arrays such as the Utah array).

What has a BCI been shown to do in people?

A widely cited 1988 study showed that scalp EEG could let people spell by attending to flashing letters, at about 2.3 characters per minute (Farwell and Donchin, 1988). In 2006 the BrainGate team reported that a man with tetraplegia used a 96-electrode array in motor cortex to control a computer cursor and open and close a prosthetic hand (Hochberg et al., Nature 2006). Since 2021, implanted BCIs have decoded attempted speech: published rates rose from a median 15.2 words per minute in 2021 (Moses et al., NEJM) to 62 and 78 words per minute in two 2023 Nature papers (Willett et al.; Metzger et al.).

Why it matters for speech BCI

Speech is the highest-bandwidth form of everyday communication, so it is the most demanding test of a BCI. People with ALS, brainstem stroke, or other causes of paralysis may keep intact speech areas in the brain while losing control of the muscles that produce speech. A BCI that reads those intact signals can, in principle, turn attempted speech back into text or voice, which is what a speech neuroprosthesis does.

Questions

Related: Speech neuroprosthesisNeural decoderEEG vs intracortical recordingElectrocorticography (ECoG)Utah arrayEndovascular BCI

Sources

1. Wolpaw et al., Brain-computer interfaces for communication and control, Clinical Neurophysiology 113, 767-791 · 2002
2. Farwell and Donchin, Talking off the top of your head: toward a mental prosthesis utilizing event-related brain potentials, Electroencephalography and Clinical Neurophysiology 70, 510-523 · 1988
3. Hochberg et al., Neuronal ensemble control of prosthetic devices by a human with tetraplegia, Nature 442, 164-171 · 2006
4. Moses et al., Neuroprosthesis for Decoding Speech in a Paralyzed Person with Anarthria, New England Journal of Medicine · 2021
5. Willett et al., A high-performance speech neuroprosthesis, Nature 620, 1031-1036 · 2023

Facts on this page are drawn from the sources listed and were accurate to the best of our knowledge on the update date shown. Company and product names are trademarks of their owners and are used for identification only. Nothing here is medical, investment or legal advice. Companies and individuals can request a correction via the corrections page.