Neuroimaging Data Model

Neuroimaging Data Model
Name	Neuroimaging Data Model
Established	2010s
Discipline	Neuroimaging, Data science

Neuroimaging Data Model The Neuroimaging Data Model is a community-driven framework for organizing, describing, and sharing brain imaging datasets and associated metadata across platforms such as Human Connectome Project, OpenNeuro, Alzheimer's Disease Neuroimaging Initiative, UK Biobank, and Human Brain Project. It provides conventions that enable reproducible analyses, integration with computational resources like Amazon Web Services, Google Cloud Platform, and institutional infrastructures such as National Institutes of Health, European Research Council, and Wellcome Trust. The model interacts with standards and tools from entities including DICOM, NIfTI, GitHub, Neuroinformatics, and consortia such as International Neuroinformatics Coordinating Facility.

Overview

The model formalizes how modalities like functional magnetic resonance imaging, diffusion tensor imaging, electroencephalography, magnetoencephalography, and positron emission tomography are represented alongside behavioral assays and clinical data drawn from projects like Alzheimer's Disease Neuroimaging Initiative, ADNI, Human Connectome Project, UK Biobank, and Human Brain Project. It emphasizes compatibility with file formats developed by standards organizations such as Digital Imaging and Communications in Medicine, NIfTI, and toolchains maintained in repositories on GitHub, enabling workflows across platforms like Neurosynth, SPM (software), FSL, AFNI, and FreeSurfer. The model supports metadata practices promoted by funders such as Wellcome Trust, National Institutes of Health, European Research Council, and journals like Nature Neuroscience and Neuron.

History and development

Origins trace to collaborative efforts linking groups behind Human Connectome Project, OpenfMRI, International Neuroinformatics Coordinating Facility, and laboratories affiliated with Massachusetts Institute of Technology, Stanford University, University of Oxford, University College London, and Harvard University. Key milestones include adoption of conventions following workshops hosted at National Institutes of Health, community proposals discussed at conferences such as Organization for Human Brain Mapping, Society for Neuroscience, and IEEE Engineering in Medicine and Biology Society. Influential contributors include teams that produced datasets for Alzheimer's Disease Neuroimaging Initiative, UK Biobank, and initiatives sponsored by Wellcome Trust and European Research Council, aligning with data policies from National Institutes of Health and editorial standards at Nature and Science.

Data model components and formats

Core components integrate imaging files in formats standardized by Digital Imaging and Communications in Medicine and NIfTI, event and behavioral logs inspired by protocols from OpenNeuro and Human Connectome Project, and metadata schemas coordinating ontologies referenced by Gene Ontology-adjacent projects and terminologies used in repositories such as NeuroLex and entities developed at International Neuroinformatics Coordinating Facility. The model prescribes directory layouts, JSON sidecar metadata following precedents from Brain Imaging Data Structure communities, and tabular phenotype descriptions compatible with formats used by Alzheimer's Disease Neuroimaging Initiative, UK Biobank, and OpenfMRI. Integration points include provenance standards exemplified by W3C PROV and identifier systems employed by ORCID, Digital Object Identifier, and institutional repositories at Massachusetts Institute of Technology and University of Oxford.

Implementation and tools

Implementations are available as software libraries, command-line utilities, and web services developed in ecosystems associated with GitHub, Python (programming language), R (programming language), and platforms like Docker and Singularity. Popular tools and pipelines include wrappers for SPM (software), FSL, AFNI, FreeSurfer, and workflow engines used by projects at Stanford University, Harvard University, Massachusetts Institute of Technology, and collaborations with cloud vendors such as Amazon Web Services and Google Cloud Platform. Community-contributed validators, converters, and visualization utilities are shared through channels such as GitHub, workshops at Organization for Human Brain Mapping, and training programs supported by Wellcome Trust and National Institutes of Health.

Applications and use cases

Use cases span large-scale population studies led by UK Biobank and Human Connectome Project, disease-focused consortia like Alzheimer's Disease Neuroimaging Initiative and clinical research at institutions including Massachusetts General Hospital, Johns Hopkins University, Mayo Clinic, and University College London. It supports meta-analyses performed with tools from Neurosynth and multicenter trials coordinated through agencies such as National Institutes of Health and European Commission. Applications include multimodal integration for projects at Human Brain Project, translational research in neurodegenerative disorders funded by Wellcome Trust, and machine learning efforts leveraging compute at Google Cloud Platform and Amazon Web Services.

Standards, interoperability, and governance

Governance relies on community consortia, working groups convened by International Neuroinformatics Coordinating Facility, platform stewards like OpenNeuro, and funding agencies including National Institutes of Health, European Research Council, and Wellcome Trust. Interoperability is achieved through alignment with standards from Digital Imaging and Communications in Medicine, metadata frameworks propagated by W3C, identifier systems like ORCID and Digital Object Identifier, and policy instruments influenced by publishers such as Nature and Science. Ongoing stewardship involves contributor agreements hosted on GitHub, community calls associated with Organization for Human Brain Mapping and governance models similar to those used by Human Connectome Project and Human Brain Project.

Category:Neuroimaging