Open neuroimaging data,
accelerating the development of innovative neurotechnologies
BayesCore NeuroLab connects you to more than 1,800 public neuroimaging datasets validated against the BIDS standard — MRI, fMRI, EEG, MEG, iEEG, PET and fNIRS — with intelligent search, direct downloads and an expert AI assistant to support every stage of your research.
Available modalities
All data follow the international BIDS standard (Brain Imaging Data Structure), ready for use in reproducible pipelines.
MRI / fMRI
Structural and functional imaging: anatomy, cognitive tasks, resting-state, diffusion (DTI/DWI) and ASL.
EEG
Scalp electroencephalography: evoked potentials, oscillations, sleep, epilepsy and BCI.
iEEG / ECoG
High-resolution intracranial recordings: functional mapping and direct cortical dynamics.
MEG
Magnetoencephalography: neural dynamics with millisecond resolution and source localization.
PET
Positron emission tomography: metabolism, receptors and molecular neuroimaging.
fNIRS
Functional near-infrared spectroscopy: portable cortical hemodynamics.
Multimodal
Studies combining EEG+fMRI, MEG+MRI and other integrations of high informational richness.
Behavior
Task events, questionnaires and phenotypic data that accompany the acquisitions.
Understand the BIDS standard and the neuroimaging modalities
📐 BIDS — Brain Imaging Data Structure
BIDS is not an acquisition technique, but an international standard for organizing neuroimaging data: it defines file names, folder structure and metadata (JSON and TSV) consistently across laboratories. This makes the data self-explanatory and processable by automated, reproducible pipelines (fMRIPrep, MRIQC, MNE-BIDS…), reducing curation errors and enabling open science at scale. Every dataset in the collection is validated against this standard.
🧲 MRI — Magnetic Resonance Imaging (structural)
It uses intense magnetic fields and radiofrequency pulses to excite the hydrogen protons of tissues and form anatomical images with high spatial resolution (typically ~1 mm). It allows volume, cortical thickness and white matter integrity to be measured (with diffusion, DTI/DWI). It is the anatomical basis on which the other analyses rest. It does not use ionizing radiation.
🌊 fMRI — functional Magnetic Resonance Imaging
It measures brain activity indirectly through the BOLD signal (blood-oxygen-level dependent): more active regions receive more oxygenated blood, altering the local magnetic signal. Excellent spatial resolution (millimeters), but temporal resolution limited (seconds) by the slowness of the hemodynamic response. Used in cognitive tasks and in the resting state (functional connectivity).
⚡ EEG — Electroencephalography
It records, through electrodes on the scalp, the electrical potentials generated by the synchronized activity of large populations of cortical neurons (above all the postsynaptic potentials of pyramidal cells). Excellent temporal resolution (milliseconds), ideal for evoked potentials, oscillations (delta to gamma) and sleep; spatial resolution limited by the conduction of the signal through the skull and tissues.
🧭 MEG — Magnetoencephalography
It detects the extremely weak magnetic fields (femtotesla) produced by neuronal currents, by means of superconducting sensors (SQUIDs) or OPM magnetometers. It combines the temporal resolution of EEG with more precise source localization, since magnetic fields are less distorted by the skull. It requires magnetically shielded rooms.
🔬 iEEG — intracranial EEG (ECoG/SEEG)
Electrical recording made with surgically implanted electrodes — over the cortex (ECoG) or in depth (SEEG) — generally in patients with refractory epilepsy during pre-surgical evaluation. It offers the highest signal-to-noise ratio and spatio-temporal resolution in human electrophysiology, making it possible to study high-frequency oscillations (HFOs) and to map functions with millimetric precision.
☢️ PET — Positron Emission Tomography
A molecular imaging technique: a radiotracer (e.g. FDG for glucose metabolism, or ligands specific to receptors and proteins such as amyloid and tau) is injected and its distribution is reconstructed by detecting the photons from positron annihilation. It quantifies biochemical processes in vivo — essential in dementias, oncology and neuropharmacology — at the cost of low temporal resolution and the use of radiation.
💡 fNIRS — functional Near-Infrared Spectroscopy
It emits near-infrared light through the scalp and measures the differential absorption by oxygenated and deoxygenated hemoglobin, estimating the cortical hemodynamic response — the same underlying phenomenon as fMRI, but with portable, silent and movement-tolerant equipment. Ideal for infants, children, naturalistic settings and populations that cannot enter the scanner; reach limited to the superficial cortex.
Why BayesCore NeuroLab?
More than a repository: a complete working environment for the researcher.
Intelligent search
Search by keyword, filter by modality and sort by date or popularity. Find the right dataset in seconds, with complete metadata: authors, tasks, subjects and size.
Flexible download
Download individual files straight from the browser or the complete dataset via Amazon S3, DataLad or openneuro-py — with ready-to-copy commands.
Dr. Áxon — your personal research assistant
Discuss study design, analysis pipelines and the interpretation of results with our AI.
Research history
All your searches and downloads are recorded in your account — pick the work up where you left off and keep your review traceable.
BIDS standard
A consistent, validated data structure, compatible with fMRIPrep, MRIQC, MNE-Python, EEGLAB, FieldTrip and the whole ecosystem of open tools.
Truly open science
Data under the CC0 license (public domain), de-identified and citable by DOI. Use them in research, teaching and the development of new technologies.
Start researching now
Create your free researcher account to unlock dataset search, downloads and the expert AI assistant.