axiom-neural / generate
sfreq: 256 Hz Alpha — Relaxed

Synthesize brainwave
data, on demand.

Physics-based multi-channel EEG synthesis with artifact simulation, real-time topomap, and PSD validation.

Brain State8 states
Clinical Presetspatient sim
Parameters
Duration 10 s
Noise level 0.15
Amplitude (μV) 80
Channels6 active
Artifact SimulationNEW
Myogenic (Jaw)
T3/T4 HF burst
Line Noise
Power grid hum
Cardio (ECG)
~1 Hz QRS bleed
Waveform Output
0.00 s 0 ch
Topomapreal-time
SNR
dB ratio
Samples
@ 256 Hz
PSD Match
Wasserstein
Quality
TSTR score
Active Artifacts
None
Power Spectral DensityWelch · 512-pt FFT
Generation Log
00:00:00Axiom Neural v2.0 — engine initialized
00:00:00FFT, Welch, pink noise, artifact engine ready

Validate + Filter Sandbox

PSD matching, band-power analysis, and interactive filtering — apply notch and bandpass filters and watch the spectrum respond.

Band Power Radar
Band Power Distribution
Filter Sandbox
Notch freq 50 Hz
High-pass (Hz) 1 Hz
Low-pass (Hz) 40 Hz
PSD Overlay — Raw vs Filtered256 Hz · Welch
Wasserstein Dist.
lower is better
TSTR Score
train-synth / test-real
KL Divergence
distribution match
Corr. Coeff.
cross-channel

EEG Corpus Registry

Open-source EEG datasets compatible with the Axiom Neural signal engine.

DatasetSubjectsChannelsDurationParadigmLicenseStatus
MNE-Python — Load Dataset
import mne from mne.datasets import eegbci raw_fnames = eegbci.load_data(1, runs=[6, 10, 14]) raw = mne.io.concatenate_raws([mne.io.read_raw_edf(f, preload=True) for f in raw_fnames]) raw.filter(1., 40.) ica = mne.preprocessing.ICA(n_components=20, random_state=42) ica.fit(raw) ica.apply(raw) df = pd.DataFrame(raw.get_data().T * 1e6, columns=raw.ch_names) df['time_s'] = raw.times df.to_csv('axiom_seed.csv', index=False)

Export Dataset

Download the current synthetic EEG as a portable file format.

Format
.CSV
Comma-separated. Compatible with Python, MATLAB, Excel.
.EDF (via MNE)
European Data Format. Native EEG standard.
.JSON
Structured JSON with metadata, channels, and signal arrays.
EDF Export via MNE-Python
import mne, pandas as pd df = pd.read_csv('axiom_export.csv') info = mne.create_info(ch_names=[c for c in df.columns if c != 'time_s'], sfreq=256, ch_types='eeg') raw = mne.io.RawArray(df.drop('time_s', axis=1).values.T * 1e-6, info) raw.export('axiom_export.edf', fmt='edf') # pip install mne pyEDFlib pandas numpy

REST API Reference

FastAPI backend endpoints for programmatic EEG synthesis.

POST /api/generate200 OK
POST /api/generate { "state": "alpha", "duration_s": 10, "sfreq": 256, "channels": ["Fp1", "Fp2", "C3", "C4"], "noise": 0.15, "amplitude_uv": 80, "artifacts": ["blink", "linenoise_50hz"] }
GET /api/presets
GET /api/presets [ { "id": "eyes_closed_male", "label": "24yo Male · Resting Eyes Closed", "state": "alpha" }, { "id": "absence_seizure", "label": "45yo Female · Absence Seizure", "state": "seizure" }, ... ]