PI: Markus Kreuzthaler
Focus: The main focus lies on the AI-based semantic modeling and interpretation of text-based clinical and research data for the creation of structured and standardized patient profiles based on international standards (SNOMED CT, ICD, LOINC, HL7 FHIR, OMOP). The employed technologies combine data-driven neural approaches, such as language models, with ontologies, knowledge bases, and rule-based methods. The resulting structured information supports primary and secondary use scenarios including clinical document retrieval, data analysis, decision-making, and quality assurance.
Networking: The team’s key collaboration partners currently include KAGes (Styrian Hospital Association), the Diagnostic and Research Center for Pathology, the clinical departments of Oncology and Cardiology, and the University Center for Acute Medicine, as well as CBmed GmbH in Graz, the Austrian Primary Care Association (APCA), and the industry partners ELGA GmbH and XUND GmbH in Vienna. In Germany, close collaborations exist with the text mining company Averbis GmbH in Freiburg, the Technical University of Munich, and the German Research Center for Artificial Intelligence (DFKI) in Berlin. In addition, the group maintains strong connections with universities in Bordeaux, Curitiba, Maastricht, Murcia, and Ljubljana, the Bern University of Applied Sciences, as well as Roche Diagnostics in Basel and Belmont.
