Transforming Clinical Data with LLMs
Fine-tuning language models for structured medical information extraction
ELMTEX presents a specialized approach to extract structured information from unstructured clinical reports using fine-tuned Large Language Models.
- Extracts critical data including patient history, diagnoses, and treatments
- Evaluates performance of LLMs of various sizes for clinical information extraction
- Develops a complete workflow with user interface for healthcare professionals
- Addresses European healthcare systems' need for better interoperability and digitalization
Why it matters: This research bridges the gap between legacy unstructured clinical data and modern healthcare information systems, enabling more efficient data processing, improved clinical decision-making, and enhanced patient care coordination.