Citing Signal Studio

Cite the version you actually ran. Software changes, and a citation without a version does not let anyone reproduce your work.

Cite the version, and archive the pipeline

Find the exact version in Help / About. Archive the pipeline file alongside your data: it records which cards ran, at which versions, with which parameters. That file is what makes the analysis reproducible, and it belongs in your supplementary material.

signalstudio.bib

download

@software{signalstudio,
  title     = {Signal Studio: a node-based workspace for EEG, sEEG and iEEG},
  author    = {{NeuraCrypt}},
  year      = {2026},
  version   = {0.9.0},
  url       = {https://eeg.studio},
  publisher = {NeuraCrypt},
  note      = {Computer software}
}

signalstudio.ris

download

TY  - COMP
TI  - Signal Studio: a node-based workspace for EEG, sEEG and iEEG
AU  - NeuraCrypt
PY  - 2026
ET  - 0.9.0
PB  - NeuraCrypt
UR  - https://eeg.studio
ER  -

apa.txt

NeuraCrypt. (2026). Signal Studio: a node-based workspace for EEG,
    sEEG and iEEG (Version 0.9.0) [Computer software].
    https://eeg.studio

The repository carries a CITATION.cff, which GitHub reads to offer a citation button, and which reference managers can import directly.

CITATION.cff

download

cff-version: 1.2.0
title: "Signal Studio: a node-based workspace for EEG, sEEG and iEEG"
type: software
authors:
  - name: "NeuraCrypt"
version: "0.9.0"
date-released: "2026-09-02"
url: "https://eeg.studio"

DOI

Signal Studio does not have a DOI yet. Archiving each release on Zenodo to mint one is on the roadmap as R-006. Until that lands, cite the version and the URL, which is what the entries above do.

Once it exists, every release will get its own DOI pointing at an archived copy, so a citation will resolve even if this site does not. That is the whole reason to do it.

In the methods section

A citation alone rarely tells a reader what you did. Name the steps and their settings, the way you would for any other method.

Preprocessing was performed in Signal Studio 0.9.0 (NeuraCrypt, 2026). Signals were re-referenced to a bipolar montage, band-pass filtered between 1 and 40 Hz with a Hamming-windowed FIR filter, and epoched from -500 to 1000 ms around stimulation onset. Trials exceeding four standard deviations of the channel variance were rejected. Each processing step was validated against MNE-Python 1.8.0 on the same recording; agreement figures are published at eeg.studio/validation. The complete pipeline file is available in the supplementary material.

That last sentence but one is worth keeping. It is the part that lets a reviewer check the claim rather than take it on trust, and the validation dashboard is public precisely so it can be cited.

Citing a card or a workflow

Cards published by other people are their work. If a community card did something substantial in your analysis, cite it separately: its marketplace page lists the author, the version and the unique identifier, which is what a citation needs.

Citing the data

If you used one of the example datasets, cite the original source rather than us. Each one names where it came from and under which licence.