The abstract is read at the publisher; the summary is JClub's.
Foundation models and large language models are transforming cancer AI by integrating heterogeneous oncology data, and may enable diagnosis, biomarker discovery, prognosis, and treatment personalization.
- Why it matters: The field needs a clear, evidence-based paradigm for using broad, mixed data sources to produce clinically meaningful outputs, while addressing gaps for real-world oncology deployment.
- What they did: This narrative review proposes the “Leave No Data Behind” paradigm and critically assesses current evidence and key challenges for clinical implementation of FMs and LLMs in oncology.
- The result: The review frames how large-scale, unstructured, label-free learning across imaging, electronic health records, and molecular profiles could support multiple clinical tasks, contingent on unresolved challenges.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.