A federated, privacy-preserving framework for assessing and monitoring data quality in distributed biomedical repositories. The framework enables quality assessment at source while sharing only aggregated, differentially private metrics with a central coordination server.
- Privacy-First: Raw data never leaves the site — only aggregated quality metrics are shared
- Open Source: Fully transparent implementation with community oversight
- Privacy Preserving: Differential privacy mechanisms protect individual records
- Easy to Deploy: Containerized deployment with automatic updates
- Extensible: Modular architecture supports multiple data sources and quality checks
Technology Readiness Level 7 - Stable enough for experimentation and pilot deployments against HL7 FHIR endpoints implementing the BBMRI.de FHIR profiles.
The Current implementation focuses on clinical data exposed via HL7 FHIR. Additional connectors for OMOP, SQL databases, and other research formats are planned based on community needs.
# Start with Docker Compose
git clone https://github.com/BBMRI-cz/data-quality-framework.git
cd data-quality-framework
docker compose up -d
# Access the agent at http://localhost:8081
# Default credentials: admin / adminpass- Issues: GitHub Issues for bug reports and feature requests
- Security: Report security vulnerabilities via GitHub Security Advisories
Licensed under the GNU General Public License v3.0
Developed by BBMRI-ERIC & Masaryk Memorial Cancer Institute