AI-Powered Structuring of Medical Data Advances Digitalization: FHIR-Starter Project

Details

The 'FHIR-Starter' project involves Insiders Technologies as a central technology partner, collaborating with the Fraunhofer Institute for Experimental Software Engineering IESE and the Berlin Institute for Health Research at Charité. Traditionally, exchanging medical information requires manually transferring data from unstructured sources, such as PDF doctor's notes, into Practice Management Systems (PVS) or Hospital Information Systems (KIS), which is time-consuming and error-prone. This project aims to solve this by developing an innovative software service that uses Artificial Intelligence (AI)—specifically Large Language Models (LLMs) and Natural Language Processing (NLP)—to automatically read and structure medical documents. The core function is the analysis of unstructured full-text documents and their conversion into standardized data structures. Crucially, these extracted information are then mapped to the FHIR format, which is the international standard for health data exchange. This enables seamless integration into existing hospital and practice IT systems. Beyond mere automation, the project emphasizes ensuring data quality and security while maintaining GDPR compliance when processing sensitive medical data. Insiders Technologies is responsible for developing the LLM, providing the core component that extracts relevant information from documents and converts it into FHIR format. This technology aims to reduce administrative burdens on healthcare professionals, allowing them to focus on patient care.