Collaborative framework on responsible AI in LLM-driven CDSS for precision oncology leveraging real-world patient data
Authors:
Open Access
- 4 December 2025
- journal article
- research article
- Published by Springer Nature in npj Precision Oncology
- Vol. 10 (1)
Abstract
Precision oncology leverages real-world data, essential for identifying biomarkers and therapies. Large language models (LLMs) can aid at structuring unstructured data, overcoming current bottlenecks in precision oncology. We propose a framework for responsible LLM integration into precision oncology, co-developed by multidisciplinary experts and supported by Cancer Core Europe. Five thematic dimensions and ten principles for practice are outlined and illustrated through application to uterine carcinosarcoma in a thought experiment.Keywords
This publication has 73 references indexed in Scilit:
- Testing and Evaluation of Health Care Applications of Large Language ModelsJAMA, 2025
- Ethics of Medical AI2024
- Creation and Adoption of Large Language Models in MedicineJAMA, 2023
- ChatGPT—A promising generative AI tool and its implications for cancer careCancer, 2023
- Explanation of machine learning models using shapley additive explanation and application for real data in hospitalComputer Methods and Programs in Biomedicine, 2022
- Challenges and opportunities beyond structured data in analysis of electronic health recordsWIREs Computational Statistics, 2021
- Predicting Drug Response and Synergy Using a Deep Learning Model of Human Cancer CellsCancer Cell, 2020
- AI4People—An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and RecommendationsMinds and Machines, 2018
- Integrated Molecular Characterization of Uterine CarcinosarcomaCancer Cell, 2017
- Rare cancers are not so rare: The rare cancer burden in EuropeEuropean Journal of Cancer, 2011