Enabling Just-in-Time Clinical Oncology Analysis With Large Language Models: Feasibility and Validation Study Using Unstructured Synthetic Data
This feasibility study demonstrated that a frontier LLM (Gemini 2.5 Pro) can successfully perform high-fidelity data extraction, multiconditional querying, and automated survival analysis directly from unstructured text. These results provide a foundational proof of concept for the JIT clinical analysis approach. However, these findings are confined to synthetic patients, and rigorous validation on real-world clinical data is an essential next step before clinical implementation can be considered.
Authors
- Peter May (ORCID: https://orcid.org/0000-0002-9045-0595)
- Christoph Seidel (ORCID: https://orcid.org/0000-0002-6023-6336)
- Julian Greß
- Markus K. Schuler (ORCID: https://orcid.org/0000-0003-0764-1049)
- Sina Nokodian
- Johannes Jung (ORCID: https://orcid.org/0000-0003-0137-7929)
- Florian Schröder
- Sebastian Sommer (ORCID: https://orcid.org/0000-0002-7446-7629)
Institutions
- Universität Hamburg (DE)
- TUM Klinikum (DE)
- University Medical Center Hamburg-Eppendorf (DE)
- Medizinisches Versorgungszentrum Prof. Mathey, Prof. Schofer (DE)
- Universitätsmedizin Göttingen (DE)
- Stiftung Oskar Helene Heim (DE)
- Munich Center for Quantum Science and Technology (DE)
Publication Details
- Journal
- JMIR Medical Informatics
- Published
- 2025-10-23
- DOI
- https://doi.org/10.2196/78332
- Citations
- 3
- Primary Topic
- Radiomics and Machine Learning in Medical Imaging
- Type
- article
- Field-Weighted Citation Impact
- 4.08