Skip to content
The BCI Briefing

ToolsHow is it built?

Tools for builders

The software, datasets and technical infrastructure that researchers and engineers use to build brain-computer interfaces. Tools are organised by the job each does in the workflow, from reading samples off a device to closing the loop, rather than as a list of links.

23
tools
10
datasets
4
standards

Open Source Radar · by role in the workflow

23 of 23 tools

  1. Hardware

  2. Acquisition

  3. Streaming

  4. Processing

  5. Decoding

  6. Applications

  7. Data & standards

  • Mature or widely used
  • Active, young or specialised
  • Quiet
  • Real-time capable

Labels are descriptive, not scores. Outlined neighbours are tools that work alongside or instead of the one in focus. Repository facts were read from GitHub on 3 Oct 2026.

Builder stacks

Reference architectures

Curated example stacks for three common goals. Swap any step for an alternative, or filter by licence, language and real-time capability to see what remains.

From a headset to a decoder driving feedback on screen, in Python. A reference architecture, not the only correct one. Swap any step to see alternatives.

  1. 01Acquisition

    BrainFlow

    Read samples from the amplifier.

    Swap

  2. 02Streaming

    Lab Streaming Layer

    Share time-stamped data across processes.

    Swap

  3. 03Preprocessing

    MNE-LSL

    Filter, re-reference and epoch online.

    Swap

  4. 04Feature extraction

    pyRiemann

    Covariances or band power.

    Swap

  5. 05Decoder

    pyRiemann

    Classify intent.

    Swap

  6. 06Feedback

    PsychoPy

    Present stimuli and results.

    Swap

Release tracking

What is moving

Read from each project's public repository. Dates are facts; the groupings are ours.

New releases

Most recent tagged releases on GitHub.

Projects to watch

Young projects under active development.

  • ezmsg

    A Python framework for building real-time signal-processing graphs with shared-memory messaging.

  • torch_brain

    A library for designing and training deep-learning models on neural population data.

Gone quiet

No recent release. Check maintenance before depending on them.

Neural data standards

Standards

Shared formats are what make a dataset usable by someone other than its author.

Explore datasets