Research and publications
Research-grade data, built by clinicians who publish.
Firemetrics comes out of clinical research. The platform is the technical backbone of a national research network, supports registries and cohort studies, and has been used in peer-reviewed work.
RACOON · Network University Medicine
The backbone of a radiology network spanning all 38 German university hospitals
RACOON, the Radiological Cooperative Network of the Network University Medicine (NUM), connects every German university hospital for multicentre imaging research. Firemetrics provides the technical backend for RACOON-CENTRAL and all nodes.
- **FHIR data space** / for study data and metadata
- **Relational FHIR engine** / for complex cross-site queries
- **PACS and object storage** / DICOM, DICOMweb, S3
- **Dashboards** / visualisation across sites
Selected on "objective performance criteria, especially for complex, cross-site queries."
Prof. Dr. Tobias Penzkofer, co-coordinator of RACOON
Portrait: Prof. Dr. med. Felix Nensa in a clinical setting
Co-founder
Prof. Dr. med. Felix Nensa
Radiologist and co-founder of Firemetrics, working at the Institute for Artificial Intelligence in Medicine at University Hospital Essen [exact title and wording to be confirmed by Prof. Nensa]. He developed the melanoma dashboard at Essen and writes regularly about FHIR, relational data models and AI agents in healthcare.
Radiology Clinical AI FHIR
Talk to his team
Publications, articles and talks
Selected work
- Real-time management of blood products
- Transfusion Medicine · authors and DOI
- Read the paper
- Article · 2026
Healthcare agents don't need search, they need precise access to clinical reality
- Read
- Article · 2025
SQL-on-FHIR vs. Firemetrics: two very different approaches to FHIR analytics
Article · 2025
FHIR as a semantic layer for business analytics
Talk · 2024
FHIR: the established standard for healthcare data interoperability
Article · 2024
The relational FHIR data model
For research teams
What researchers use it for
Cohorts in seconds
At the University Hospital Essen skin cancer centre, complex cohort queries that used to need manual Excel work run in seconds. 3,949 melanoma patients in one view.
Registries and mandatory reporting
Structured documentation filled from source documents, e.g. 48 TAVI fields per patient for registry and quality reporting.
Multicentre research
Research networks query structured data across sites, with the same foundation at every site, as in the RACOON network.
"Direct database access saves many intermediate steps and significantly accelerates the entire development process."
Academic partners
Where we work with academia
RACOON Network University Medicine (NUM) University Hospital Essen
Names shown as published on the Firemetrics blog; confirm for the website.
Planning a registry or cohort study?
Talk to the team that builds research on hospital data every day.