Narrative-based explainable AI for precision oncology survival prediction
Bioinformatics · 30.10.2026
DOI: 10.1093/bioinformatics/btag806 ↗Kurz gefasst
Fachartikel in der Zeitschrift Bioinformatics, online erschienen am 30.10.2026. Verfasst von Ziyun Pan und 7 weiteren Autor:innen. Beteiligte Einrichtung: AI for Genome Interpretation lab (AI4GI), Institut de Génétique Moléculaire de Montpellier (CNRS), Université de Montpellier, 34090, Montpellier, Fra…. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Motivation Explainable and accurate Survival Prediction (SP) remain an open challenge in Precision Oncology. Early SP methods achieved interpretability through simple models, often at the expenses of prediction performance. More recent Machine Learning (ML) methods can learn complex patterns, but they often resemble inscrutable black boxes. Here we propose SurvXAI, a new framework that combines reliable SP with the latest eXplainable AI (XAI) approach. Results We first train a Random Survival Forest model to predict the survival function of each patient. We then interpret these prediction by extracting patient-specific feature contributions with SurvSHAP, and we feed them into locally deployed Large Language Models (LLMs), to translate the numeric individual-level interpretation of the SP into natural language narrative explanations. Since conventional numerical breakdowns of the features contributions to the final predictions are often perceived as un-intuitive to non-ML experts, SurvXAI overcomes this limitation by automatically translating them into narrative explanations that can be readily used by clinicians, to facilitate the interpretation of model predictions. We validated our SP pipeline on six oncological datasets, including an in-house Multiple Myeloma cohort, comparing it with nine state of the art SP predictors. We then quantified the quality and usefulness of SurvXAI narrative explanations with a blind evaluation carried by seven human experts. This evaluation showed that larger LLMs produced more coherent explanations with accurate feature ranking and clinically plausible reasoning. SurvXAI demonstrates the potential of integrating explainable SP with LLMs to enable trustworthy, personalized interpretation in oncology and other clinical fields. Availability and Implementation SurvXAI is available at https://github.com/AI-for-Genome-Inter pretation-lab/SurvXAI_pipeline. The module for narrative generation is available at https://github.com/AI-for-Genome-Interpretation-lab/SurvX AI_stories.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- AI for Genome Interpretation lab (AI4GI), Institut de Génétique Moléculaire de Montpellier (CNRS), Université de Montpellier, 34090, Montpellier, France.
- Department of Medical Sciences, University of Torino, 10123 Torino, Italy.
- Institut de Génétique Humaine (IGH), CNRS UMR 9002, Université de Montpellier, 34396, Montpellier, France.
- Institut de Modélisation Mathématique de Montpellier, Université Montpellier, 34095, Montpellier, France.
- Department of Biological Hematology, The University Hospital of Montpellier, 34295, Montpellier, France.
Zitieren
APA
Pan, Z., Codicè, F., Bret, C., Garcia de Paco, E., Thomas, M., Mas, A., Moreaux, J., & Raimondi, D. (2026). Narrative-based explainable AI for precision oncology survival prediction. Bioinformatics. https://doi.org/10.1093/bioinformatics/btag806
Vancouver
Pan Z, Codicè F, Bret C, Garcia de Paco E, Thomas M, Mas A, et al. Narrative-based explainable AI for precision oncology survival prediction. Bioinformatics. 2026. doi:10.1093/bioinformatics/btag806
BibTeX
@article{Pan2026,
author = {Pan, Ziyun and Codicè, Francesco and Bret, Caroline and Garcia de Paco, Elvira and Thomas, Morgane and Mas, André and Moreaux, Jerome and Raimondi, Daniele},
title = {Narrative-based explainable AI for precision oncology survival prediction},
journal = {Bioinformatics},
year = {2026},
doi = {10.1093/bioinformatics/btag806}
}
EUVIMED