跳到正文
原文
Nature 健康护理主题·· 23 天前AI 评分31

带偏置的 FedSDS 联邦学习策略用于非 IID 多疾病生存预测研究

Leveraging biased FedSDS as a federated learning strategy for research on non-IID multi-disorder studies

AI 导读

联邦学习策略 FedSDS 通过在各客户端本地生成合成数据并共享,避免传输真实敏感数据,在骨髓增生异常综合征、低级别胶质瘤和乳腺癌三项基于 DeepSurv 的生存预测任务中,非 IID 设置下性能优于 FedAvg 及其他参数优化策略。

正文

Abstract

Federated learning is a powerful approach to overcome privacy and data sharing issues, especially in small sample healthcare settings. However, parameter aggregation strategies, such as FedAvg, are prone to fail in case of non-IID data distributions, which often arise in real-world scenarios. In fact, data provided by different institutions can exhibit systematic heterogeneity across cohorts, such as age, genotype, and life expectancy, reflecting institution-specific biases. Federated Synthetic Data Sharing (FedSDS) represents a valuable solution to the problem of non-IID data. It is a federated learning strategy based on locally generating synthetic datasets from real data and sharing them between the federated clients, avoiding the transfer of real sensitive data. FedSDS is applied to a survival prediction task, performed using a DeepSurv model architecture in 3 different healthcare use cases: Myelodysplastic Syndrome, Low Grade Glioma, and Breast Cancer. We demonstrate that this approach outperforms FedAvg and other strategies based on model parameters optimization in non-IID settings. In particular, the possibility to locally bias shared synthetic data allows preserving the local data distribution of each participating client and significantly improving the performances of federated models. This study shows that biased FedSDS is effective in three different healthcare use cases, such as Myelodysplastic Syndromes, Low Grade Glioma, and Breast Cancer. Our findings suggest that FedSDS may have potential relevance for healthcare applications, as it can support the selection and biasing of samples for dataset augmentation and improve the performance of federated models in non-IID multi-disorder settings. The potential use of this approach for clinical trial cohort selection, however, would require further investigation and dedicated validation.

Abbreviations

AI:

Artificial intelligence

BRCA:

Breast invasive carcinoma

CDF:

Cumulative distribution function

CV:

Cross-validation

DL:

Deep learning

DP:

Differential privacy

EHR:

Electronic health records

FedAvg:

Federated averaging

FedNova:

Federated normalized averaging

FedProx:

Federated proximal

FedSDS:

Federated synthetic data sharing

FL:

Federated learning

GBSG:

German Breast Cancer Study Group

IID:

Independent and identically distributed

KDE:

Kernel density estimate

KM:

Kaplan–Meier

LGG:

Low grade glioma

METABRIC:

Molecular Taxonomy of Breast Cancer International Consortium

MDS:

Myelodysplastic syndrome

ML:

Machine learning

TCGA:

The cancer genome atlas

VAE:

Variational AutoEncoder

VAE-BGM:

Variational AutoEncoder with Bayesian Gaussian Mixture

Funding

This research is supported by European Union–Horizon 2020/2023 program and Innovative Health initiative, IHI (GenoMed4All project #101017549 to G.C.; Synthema project, #1101095530 to G.C.; SYNTHIA project #101172872 to G.C). AIRC Foundation (Associazione Italiana per la Ricerca contro il Cancro, Milan Italy–Project #26216 to G.C.; PRIN (Ministry of University & Research, Italy–Project 20229B28PE to G.C). SYNTHIA (Synthetic Data Generation framework for integrated validation of use cases and AI healthcare applications) is supported by the Innovative Health Initiative Joint Undertaking (IHI JU) under grant agreement No 101172872. The JU receives support from the European Union’s Horizon Europe research and innovation programme, COCIR, EFPIA, Europa Bío, MedTech Europe, Vaccines Europe and DNV. The UK consortium partner, The National Institute for Health and Care Excellence (NICE) is supported by UKRI Grant 10132181. Funded by the European Union, the private members, and those contributing partners of the IHI JU. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the aforementioned parties. Neither of the aforementioned parties can be held responsible for them.

Author information

Author notes

  1. Juan Parras, Enrico Giampieri and Gastone Castellani have contributed equally to this work.

Authors and Affiliations

  1. IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy

    Francesco Casadei

  2. Information Processing and Telecommunications Center ETS Ingenieros de Telecomunicación, Universidad Politécnica de Madrid, Madrid, Spain

    Patricia A. Apellániz & Juan Parras

  3. Department of Medical and Surgical Sciences, University of Bologna, Bologna, Italy

    Luciana Carota, Enrico Giampieri & Gastone Castellani

  4. IRCCS Humanitas Research Hospital, Rozzano, Milan, Italy

    Gianluca Asti & Matteo G. Della Porta

  5. Department of Biomedical Sciences, Humanitas University, Pieve Emanuele, Milan, Italy

    Matteo G. Della Porta

  6. IRCCS Azienda Ospedaliero - Universitaria di Bologna, Bologna, Italy

    Enrico Giampieri & Gastone Castellani

Authors

  1. Francesco Casadei
  2. Patricia A. Apellániz
  3. Luciana Carota
  4. Gianluca Asti
  5. Matteo G. Della Porta
  6. Juan Parras
  7. Enrico Giampieri
  8. Gastone Castellani

Corresponding author

Correspondence to Francesco Casadei.

Ethics declarations

Competing interests

The authors declare that they have no competing interests.

Ethics approval and consent to participate

The original MDS cohort includes a retrospective dataset of 2043 samples provided by the GenoMed4All and Synthema consortia, together with a publicly available cohort of 2384 samples provided by the International Working Group for the Prognosis of MDS (IWG-PM), resulting in a total of 4427 patients. The original collection of the non-public cohort was approved by the Humanitas Research Hospital Ethics Committee (approval date: 2021-01-27, approval number: 2782). Written informed consent was obtained from each participant. All procedures were conducted in accordance with the ethical standards of the relevant institutional and/or national research committees and with the 1964 Declaration of Helsinki and its subsequent amendments. The source study was conducted by the EuroMDS consortium and registered at ClinicalTrials.gov (NCT04889729 on 2021-05-12). The Cancer Genome Atlas Low Grade Glioma (LGG) is a publicly available, de-identified data collection used to build a research community focused on connecting cancer phenotype with genotype33. The TCGA-Breast Invasive Carcinoma (BRCA) is a publicly available, de-identified data collection aimed at fulfilling the same objectives of TCGA-LGG dataset34.

Additional information

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Casadei, F., A. Apellániz, P., Carota, L. et al. Leveraging biased FedSDS as a federated learning strategy for research on non-IID multi-disorder studies. Sci Rep (2026). https://doi.org/10.1038/s41598-026-76112-9

Download citation

  • Received: 23 May 2026

  • Accepted: 12 October 2026

  • Published: 28 October 2026

  • DOI: https://doi.org/10.1038/s41598-026-76112-9

Keywords

来源:Nature 健康护理主题 · nature.com