- Klinik
- Anästhesie
- Intensivmedizin
- Notfallmedizin
- Schmerzmedizin
- Weiterbildung/Lehre
- Forschung
- Niedergelassene
- Kontakt
The Medical Informatics Research Group is an academic research group at University Medicine Greifswald, which comprises the Medical Faculty and University Hospital of the University of Greifswald.
The Medical Informatics research group develops, implements, and evaluates digital methods that make clinical knowledge and routinely collected healthcare data usable for concrete decisions in patient care.
We combine medical informatics, clinical expertise, and software development. Our work ranges from computer-interpretable clinical guidelines and interoperable data models to automated quality monitoring and predictive AI models.
Digital Clinical Guidelines & Knowledge-Based Decision Support
We transform guideline recommendations and quality indicators into standardized, computer-interpretable formats and link them with routinely collected clinical data. This enables automated assessment of the applicability and implementation of evidence-based recommendations and supports patient-specific clinical decision-making.
Artificial Intelligence in Anesthesiology and Intensive Care Medicine
We develop and evaluate models for predicting clinically relevant events from longitudinal routine data. Our work focuses on foundation models, clinically relevant validation, and generalizability across hospitals.
Digital Quality Benchmarking & Interoperability
We standardize intensive care routine data, develop reusable clinical logic, and enable comparable quality measurements across institutions. To this end, we combine approaches including OMOP, FHIR, international terminologies, and Clinical Quality Language (CQL).
August 2026: Benchmark for Guideline-Based Medical Question Answering
With cpgQA-DE, we developed and published a validated German-language benchmark for evaluating the ability of Large Language Models to answer questions based on German clinical practice guidelines.
March 2026: Clinically Realistic Evaluation of Sepsis Prediction Models
We show how strongly measured model performance depends on the chosen evaluation strategy and the intended clinical use case. The results highlight the importance of external validation and clinically realistic evaluation endpoints.
January 2026: Multicenter Evaluation of an Automated Guideline Adherence Monitoring System
We developed an interoperable system for automatically assessing guideline applicability and adherence, applied it to more than 82,000 ICU stays and approximately 2.2 million patient-days across five university hospitals, and validated it against physician assessments.
Contact: {first.last}@med.uni-greifswald.de
We offer Bachelor’s, Master’s, and doctoral research projects in interdisciplinary projects with direct clinical relevance. Interested students are very welcome to get in touch.
