Skip to content
The BCI Briefing

TechnologyHow do neural interfaces work?

How we interface with the nervous system

Every neural interface trades access, resolution and invasiveness against each other. This explorer places 32 of them on one map, by what they do and how far they have come, so that an EEG cap, a deep brain stimulator and an intracortical array can be read on the same terms.

BCI is evolving from decoding the nervous system to interacting with it.

  1. Reading
  2. Writing
  3. Closed loop
  4. Biohybrid

32 interfaces

Read

Bidirectional

Write

Non-invasive

Around the head

On the skin

Minimally invasive

Vessel or skull window

Invasive

Nerve or sense organ

On the cortex

In the cortex

Deep brain

Biohybrid · Adjacent frontier

Living neural tissue outside the body. Not an interface with a person.

Fig. 01 — The interface design spacePosition shows function and access. Fill shows maturity. Select an interface.

Colour · function

  • Read
  • Write
  • Bidirectional
  • Biohybrid

Fill · maturity as a clinical modality

  • Established
  • Human feasibility
  • Early human
  • Preclinical
  • Research frontier
  • Adjacent frontier

ReadNon-invasiveElectrical

Neuromotor interfaces

Sensors on the skin record the electrical activity of muscles, and with it the output of the spinal motor neurons that drive them. A wrist band can turn intended hand movements into computer input.

Showing Neuromotor interfaces: Read, Non-invasive.

Where it acts
On the skin over muscle, typically at the wrist or forearm
Information flow
Muscle → skin sensors
Motor nerveForearm muscleSensor band
Fig. 02 — Where it acts, and which way information flowsSchematic, not to scale. Not a specific device.

How far it has come

As a clinical modality
EstablishedElectromyography is a routine clinical test, and myoelectric prostheses are established products.
As a BCI application
EstablishedOn sale in wristbands and prosthetic hands. It reads muscle, not brain, and needs an intact pathway to the limb.
Access
Non-invasive. Implanted muscle sensors also exist, for prosthesis control.
Mechanism
Electrical
Target
Muscle / neuromotor
Signal
Motor-unit action potentials

Characteristics · read interfaces

Spatial resolutionMuscles, down to single motor units
Temporal resolutionMilliseconds
Signal richnessMotor output only, at high fidelity
CoverageOne muscle group
InvasivenessNone
Long-term stabilityStable; sensitive to where the band sits

Each track runs from low to high. A wide bar means the characteristic varies across implementations.

Strengths

  • Strong signal with no surgery
  • Wrist sEMG generalises across people without calibration

Trade-offs

  • Depends on an intact peripheral pathway
  • Reads movement commands only

Common applications

  • Computer and device input
  • Prosthetic hand control

Human evidence

Wrist sEMG models trained on thousands of participants decode handwriting at 20.9 words per minute with no per-user calibration (2025). Myoelectric prostheses add long clinical experience.

What needs to happen next

Evidence in people with weak or atypical muscle signals, the group with most to gain.

Used by
Meta Reality LabsBrainCo
Research
A generic non-invasive neuromotor interface for human-computer interaction
Tools
OpenBCI GUIBrainFlowLab Streaming LayerpyRiemann
Datasets
emg2qwerty
Use cases
ControlAugment
Patent activity
220 patent families name this interface, most often on peripheral interfaces, neural decoding, signal processing. The patent landscape

Descriptions, ranges and maturity levels are editorial analysis, kept qualitative where the literature does not support exact comparison across technologies. Specific devices differ.

All 32 interfaces as a table
Every interface, with its function, access, mechanism, target and maturity
InterfaceFunctionAccessMechanismTargetClinical modalityBCI application
ReadNon-invasiveElectricalCortexEstablishedHuman feasibility
ReadNon-invasiveMagneticCortexEstablishedResearch frontier
ReadNon-invasiveOpticalCortexHuman feasibilityResearch frontier
ReadNon-invasiveElectricalMuscle / neuromotorEstablishedEstablished
ReadMinimally invasiveElectricalCortexHuman feasibilityHuman feasibility
ReadMinimally invasiveAcousticCortex, Deep brainEarly humanPreclinical
ReadInvasiveElectricalPeripheral nerve, Muscle / neuromotorEarly humanEarly human
ReadInvasiveElectricalCortexEstablishedHuman feasibility
ReadInvasiveElectricalCortexHuman feasibilityHuman feasibility
ReadInvasiveElectricalDeep brain, CortexEstablishedEarly human
WriteNon-invasiveMagneticCortexEstablishedResearch frontier
WriteNon-invasiveElectricalCortexHuman feasibilityResearch frontier
WriteNon-invasiveAcousticCortex, Deep brainEarly humanResearch frontier
WriteInvasiveElectricalPeripheral nerveEstablishedEarly human
WriteInvasiveElectricalRetinaHuman feasibilityHuman feasibility
WriteInvasiveElectricalCochleaEstablishedEstablished
WriteInvasiveElectricalBrainstemEstablishedEstablished
WriteInvasiveElectricalCortexEstablishedEarly human
WriteInvasiveElectricalCortexEarly humanEarly human
WriteInvasiveOptical, MolecularCortex, Deep brain, RetinaEarly humanPreclinical
WriteInvasiveElectricalDeep brainEstablishedEarly human
BidirectionalInvasiveElectricalPeripheral nerve, Muscle / neuromotorEarly humanEarly human
BidirectionalInvasiveElectricalCortexEstablishedEarly human
BidirectionalInvasiveElectricalCortex, Deep brainEstablishedHuman feasibility
BidirectionalInvasiveElectricalCortex, Spinal cordEarly humanEarly human
BidirectionalInvasiveElectricalCortexEarly humanEarly human
BidirectionalInvasiveElectricalDeep brain, CortexEstablishedEarly human
BidirectionalInvasiveElectricalDeep brainEstablishedHuman feasibility
BiohybridEx vivoElectrical, BiologicalCultured neural tissueAdjacent frontierAdjacent frontier
BiohybridEx vivoElectrical, Optical, BiologicalCultured neural tissueAdjacent frontierAdjacent frontier
BiohybridEx vivoBiological, ElectricalCultured neural tissueAdjacent frontierAdjacent frontier
BiohybridEx vivoBiological, ElectricalCultured neural tissueAdjacent frontierAdjacent frontier

Use-case explorer

Start from what people want to do

Nine human goals. Each one connects the technology, the companies, the research, the studies and the tools behind it, and says plainly what is available today.

Decoding attempted speech or writing into text and voice.

Can someone who cannot speak hold a conversation again?

Research participants with paralysis have used implanted BCIs to produce text and synthesised voice from attempted speech. All such systems remain in early feasibility studies.

Latest milestoneA single surface implant decoded speech and upper-body gestures at the same time in three participants.

Technology
EEGMEG / OPMEndovascularECoGIntracorticalsEEG
Companies
NeuralinkSynchronPrecision NeuroscienceParadromicsBlackrock NeurotechBrainGateStairMedNeuroXessNeuCyber NeuroTechABILITY NeurotechJiMEDg.tec medical engineeringArayaCognixion
Research
Simultaneous speech and gesture decoding for multimodal communication in paralysisAn Accurate and Rapidly Calibrating Speech NeuroprosthesisAn instantaneous voice-synthesis neuroprosthesisA high-performance speech neuroprosthesisA high-performance neuroprosthesis for speech decoding and avatar control+10 more
Clinical
VOICEConnect-OneBrainGate2BRAVOUNPCortiCom+3 more
Tools
BRANDtorch_brainimg_pipeMNE-PythonBraindecode
Datasets
Brain-to-Text '25Brain-to-Text Benchmark '24LibriBrain / LibriBrain100Podcast ECoG
Frontier
Speech & CommunicationCognitive Interfaces
On the calendar
BCI Meeting 2027

Why do such different devices exist for the same goal? Each trades signal quality against surgical access. Read the dossier