TRUST Principles
The Envision Portal is an open-source platform designed to streamline the management and sharing of eye imaging data. It helps researchers manage, curate, and share their data following the FAIR principles so that datasets are ready for AI-driven analysis and collaborative studies.
This page describes how Envision Portal aligns with the TRUST Principles for Digital Repositories (Transparency, Responsibility, User Focus, Sustainability, and Technology) and points to public evidence of our implementation.
Transparency
We are committed to being open about how the Envision Portal is operated, governed, and developed, so that users understand what the platform does and how their data are handled.
Our implementation:
- Open platform information: The About page describes the Portal's purpose, designated community, key features, technology stack, and current development status.
- Policies & data governance: Our Terms of Use and Privacy Policy explain conditions of use, data protection, and responsibilities for users and operators.
- Repository documentation: The Envision Portal documentation describes the repository's architecture, dataset workflows, and operational considerations, and will expand as the platform matures.
- Service status: Service uptime, response times, and planned maintenance are monitored and reported via our public status page, powered by Uptime Kuma, at status.envisionportal.org .
Responsibility
We take responsibility for the stewardship of eye imaging datasets and associated metadata, including their quality, preservation, and appropriate use.
Our implementation:
- FAIR-aligned curation & standards: Envision Portal is built around the FAIR principles and supports community standards for clinical and imaging data (e.g., DICOM, OMOP, and structured clinical dataset models). Public examples of how we document complex datasets are available in metadata repositories such as our metadata file examples .
- Preservation & backups: We maintain daily snapshots of our database and store backups in Azure-backed object storage to support long-term preservation and disaster recovery. As the platform grows, we will continue to refine integrity monitoring and fixity checks.
- Access controls & responsible reuse: Public and controlled datasets are clearly distinguished. For controlled datasets, access requests capture the requester’s name, affiliation, email, and reason for access, and are handled according to dataset-specific conditions and project-level governance.
- Dataset-level documentation: Each dataset overview page describes key metadata, access conditions, and links to external documentation where available, so that users understand how data may be used responsibly.
User Focus
Our services are designed around the needs of our designated community: ophthalmology and vision science researchers, AI/ML developers, and data stewards who work with eye imaging datasets.
Our implementation:
- Public dataset catalog: Users can discover and explore published datasets via the datasets catalog . Each dataset overview page provides key metadata, access conditions, and a metrics section with usage indicators (e.g., views and downloads).
- Guided submission workflows: For invited contributors, Envision Portal offers guided workflows that support data standardization, de-identification, metadata completeness, and appropriate access controls, making it easier to share data that are responsibly reusable by the community.
- Support & communication: General questions and access-related inquiries can be submitted via our Contact page . Bugs and feature requests can be filed as GitHub issues at github.com/EyeACT/envision-portal/issues .
- Iterative improvement: User feedback informs ongoing improvements to submission workflows, dataset presentation, and supporting documentation.
Sustainability
We work with institutional and funding partners to ensure that Envision Portal remains a viable, well-governed service that can support long-term access to eye imaging datasets.
Our implementation:
- Institutional & project partners: Envision Portal is developed by the FAIR Data Innovations Hub team as part of the Eye ACT study, in collaboration with the University of Washington and Kaiser Permanente Washington.
- Funding: Our contribution to Eye ACT is funded through a subaward from the National Institute on Aging (NIA) grant R01AG060942. Funding acknowledgements are included in project documentation and related materials.
- Long-term planning: Long-term access plans for datasets hosted in Envision Portal are developed in collaboration with our funding and institutional partners. As the platform matures, we will publish additional details on governance, risk management, and business continuity.
- Scalable architecture: The platform uses a modern, cloud-native architecture (Nuxt.js, Vue.js, TypeScript, PostgreSQL, and Azure Data Lake) that can scale with growing data volumes and user demand.
Technology
We employ appropriate, standards-based technologies and security practices to ensure that Envision Portal is reliable, secure, and interoperable with the broader research ecosystem.
Our implementation:
- Open-source, modern stack: Envision Portal is built with Nuxt.js, Vue.js, and TypeScript, backed by PostgreSQL and Azure Data Lake for scalable storage. The platform is open-source and designed for extensibility, with source code hosted at github.com/EyeACT/envision-portal .
- Security & access control: All traffic is served over HTTPS/TLS. Authentication and role-based access control protect non-public datasets, and Azure-managed encryption at rest is used for databases and storage. Systems are regularly updated with security patches.
- Identifiers & metadata standards: We use globally recognized research identifiers (such as DOIs, ORCID IDs, and ROR identifiers where applicable) and adopt community standards for data formats and metadata (for example, DICOM for imaging and OMOP-based mappings for clinical data).
- Documentation-first approach: Technical architecture and repository operations are documented at docs.envision.io , with further details to be added as additional functionality is introduced.
TRUST Principles – Compliance Self-Assessment
The table below provides a high-level self-assessment of how Envision Portal currently aligns with the TRUST Principles for Digital Repositories. It is intended as a living overview that will be updated as the platform evolves.
| Principle | Status | Implementation Details |
|---|---|---|
| Transparency: Repository policies, scope, and operations are clearly described and publicly accessible. | Platform purpose, scope, and current status are documented on the About page; Terms of Use and Privacy Policy are public; documentation is available at docs.envision.io; and service status is reported via a public status page. | |
| Responsibility: Repository assumes responsibility for long-term stewardship of datasets and associated metadata. | FAIR-aligned curation practices, daily database snapshots and backups in Azure object storage, clear separation of public vs. controlled datasets, and dataset-level documentation support responsible reuse. | |
| User Focus: Repository services are designed around the needs of the designated community. | The Portal targets ophthalmology and vision science researchers, AI/ML developers, and data stewards, with a public dataset catalog, guided submission workflows for invited contributors, metrics on dataset overview pages, and multiple support channels (contact form and GitHub issues). | |
| Sustainability: Repository has a sustainable organizational and funding context and plans for long-term operation. | Envision Portal is developed as part of the Eye ACT study by the FAIR Data Innovations Hub team in collaboration with the University of Washington and Kaiser Permanente Washington, supported by a subaward from NIA grant R01AG060942, using a cloud-native architecture designed for growth and long-term use. | |
| Technology: Repository uses appropriate, standards-based technologies and security practices. | The platform is built on a modern, open-source stack (Nuxt/Vue/TypeScript, PostgreSQL, Azure Data Lake) with HTTPS/TLS, role-based access control, encryption at rest, regular security updates, and use of community standards for identifiers and data formats where applicable. |
References
TRUST Principles definition as referenced from: Lin, D., Crabtree, J., Dillo, I. et al. The TRUST Principles for digital repositories. Sci Data 7, 144 (2020). https://doi.org/10.1038/s41597-020-0486-7
Questions about our implementation of the TRUST Principles?
We welcome questions and feedback about how Envision Portal implements the TRUST Principles. This page and our practices will evolve as the platform grows and new features are introduced.