EUVIMEDEUVIMED entdecken
Open Access

Linear Federated Learning for Outcome Prediction With Application to Hepatocellular Carcinoma Radiotherapy

Keyur D Shah iD, Harald Paganetti, Pablo Yepes, Theodore S Hong iD, Jennifer Y Wo, J Hannah Roberts, Eugene J Koay iD, Christian V Guthier, Ibrahim Chamseddine iD

JCO Clinical Cancer Informatics · 2025-07 · (9)

DOI: 10.1200/cci-25-00074 ↗

Kurz gefasst

Fachartikel in der Zeitschrift JCO Clinical Cancer Informatics, 2025-07. Verfasst von Keyur D Shah und 8 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiation Oncology, Winship Cancer Institute, Emory University, Atlanta, GA. Hauptthemen laut MEDLINE: Carcinoma, Hepatocellular, Liver Neoplasms. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 35 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NCI NIH HHS.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

PURPOSE Federated learning (FL) enables multi-institutional predictive modeling without sharing raw patient data, preserving privacy while leveraging diverse data sets. This study evaluates the use of linear FL (LFL) as an interpretable approach to enhance sample size and generalizability in outcome prediction. As a proof of concept, we applied LFL to patients with hepatocellular carcinoma (HCC) undergoing external beam radiotherapy (EBRT), predicting hepatic toxicity and 1-year survival (SRVy1). METHODS Patient data from Massachusetts General Hospital (MGH) and Brigham and Women's Hospital (BWH) were used to train models, whereas an independent validation data set from MD Anderson Cancer Center assessed generalizability. Logistic regression was developed to predict hepatic toxicity and SRVy1 using key clinical features, including baseline albumin, bilirubin, Child-Pugh score, liver size, and mean liver dose. The LFL approach allowed each institution to train models locally without sharing raw patient data. Model performance was evaluated using the AUC and compared between the LFL model and institution-specific models. RESULTS For survival prediction, single-institution models were limited, with AUC = 0.55-0.63, with LFL increasing it to 0.67. For toxicity prediction, external validation showed AUC = 0.68 for the MGH model and 0.69 for the BWH model, with LFL maintaining the AUC at 0.7. The model coefficients were moderate in the LFL compared with the single-institution models, indicating an ability to mitigate bias, which was also reflected by better performance on the validation data set. CONCLUSION LFL maintained or improved predictive performance over single-institution models for survival and hepatic toxicity in patients with HCC treated with EBRT while preserving model interpretability and patient privacy. These findings support LFL's role in multi-institutional collaborations.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Carcinoma, HepatocellularLiver NeoplasmsHumansFemaleMaleMiddle AgedAgedPrognosisTreatment OutcomeFederated Learning

Autor:innen und Institutionen

  • Department of Radiation Oncology, Winship Cancer Institute, Emory University, Atlanta, GA.
  • Department of Radiation Oncology, Mass General Brigham, Harvard Medical School, Boston, MA.
  • Department of Physics and Astronomy, Rice University, Houston, TX.
  • Department of Radiation Oncology, University of Texas, MD Anderson Cancer Center, Houston, TX.

ORCID-Profile: Keyur D Shah, Theodore S Hong, Eugene J Koay, Ibrahim Chamseddine

Förderung

  • NCI NIH HHS

Zitieren

APA

Shah, K. D., Paganetti, H., Yepes, P., Hong, T. S., Wo, J. Y., Roberts, J. H., Koay, E. J., Guthier, C. V., & Chamseddine, I. (2025). Linear Federated Learning for Outcome Prediction With Application to Hepatocellular Carcinoma Radiotherapy. JCO Clinical Cancer Informatics. https://doi.org/10.1200/cci-25-00074

Vancouver

Shah KD, Paganetti H, Yepes P, Hong TS, Wo JY, Roberts JH, et al. Linear Federated Learning for Outcome Prediction With Application to Hepatocellular Carcinoma Radiotherapy. JCO Clinical Cancer Informatics. 2025. doi:10.1200/cci-25-00074

BibTeX
@article{Shah2025,
  author = {Shah, Keyur D and Paganetti, Harald and Yepes, Pablo and Hong, Theodore S and Wo, Jennifer Y and Roberts, J Hannah and Koay, Eugene J and Guthier, Christian V and Chamseddine, Ibrahim},
  title = {Linear Federated Learning for Outcome Prediction With Application to Hepatocellular Carcinoma Radiotherapy},
  journal = {JCO Clinical Cancer Informatics},
  year = {2025},
  number = {9},
  doi = {10.1200/cci-25-00074}
}
Kostenlos starten →