OpenNeuro
OpenNeuro is a free, open archive of human brain-imaging datasets run by Stanford University's Center for Reproducible Neuroscience, led by Russell Poldrack, and funded as a BRAIN Initiative data archive. It launched in 2017 as the successor to OpenfMRI and now holds a large, growing collection of studies spanning several imaging methods: structural and functional MRI, diffusion MRI, PET, MEG, EEG and intracranial EEG. Every dataset is stored in the standardised Brain Imaging Data Structure (BIDS) layout, and uploads must pass BIDS validation, so files, metadata and folder conventions stay consistent across studies. Each dataset carries a version and a citable DOI.
You access the data without any account or application. Datasets can be browsed and downloaded through the web portal, pulled in bulk from the command line, cloned as git repositories (compatible with DataLad), or queried through a GraphQL API that exposes dataset metadata. At the 2021 resource paper the archive reported more than 600 datasets from roughly 21,000 participants, and it has expanded considerably since.
For RAG and AI work, OpenNeuro is a source of raw neuroimaging data rather than text, so it suits multimodal and scientific pipelines, the training and benchmarking of brain-imaging models, and retrieval over the structured metadata (tasks, participants, acquisition parameters) that accompanies each study. It is best for reproducible neuroscience and medical-imaging projects that need large, well-documented, standards-compliant datasets.
Datasets are released by default under a CC0 public-domain dedication, which places no restrictions on reuse and requires no attribution, making OpenNeuro unusually safe to build on commercially. The important caveat is ethical rather than legal: these are human-subjects neuroimaging data, so downstream use still carries responsibilities around privacy, consent and defacing, even though the licence itself is unrestricted. Within our directory it offers the open imaging counterpart to our text-based biomedical repositories.
Related sources
AlphaFold Protein Structure Database
EMBL-EBI and Google DeepMind's open database of AI-predicted protein structures, covering over 200 million sequences across almost all of UniProt. Each model ships with per-residue confidence and predicted aligned error scores, and is reachable by website, API, FTP and Google Cloud bulk access. Released under CC BY 4.0 and designed to pair with UniProt and the experimental structures in the PDB.
ClinicalTrials.gov
A US National Institutes of Health database of privately and publicly funded clinical studies. It holds structured records on more than 450,000 studies, including protocols, conditions, interventions, outcomes, and results, all in the public domain.
DrugBank
A freely accessible database that combines detailed drug data with drug target information, covering more than 15,000 drug entries. It links chemistry, pharmacology, and biology in one place, which makes it a strong knowledge source for medical and pharmaceutical RAG.
MIMIC-III / MIMIC-IV
Freely accessible critical care databases with de-identified health data from more than 40,000 intensive care patients at Beth Israel Deaconess Medical Center. They include vital signs, lab results, medications, and clinical notes. Access needs credentialing and a data-use agreement.