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
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。
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
Juan Parras, Enrico Giampieri and Gastone Castellani have contributed equally to this work.
Authors and Affiliations
IRCCS Istituto delle Scienze Neurologiche di Bologna, Bologna, Italy
Francesco Casadei
Information Processing and Telecommunications Center ETS Ingenieros de Telecomunicación, Universidad Politécnica de Madrid, Madrid, Spain
Patricia A. Apellániz & Juan Parras
Department of Medical and Surgical Sciences, University of Bologna, Bologna, Italy
Luciana Carota, Enrico Giampieri & Gastone Castellani
IRCCS Humanitas Research Hospital, Rozzano, Milan, Italy
Gianluca Asti & Matteo G. Della Porta
Department of Biomedical Sciences, Humanitas University, Pieve Emanuele, Milan, Italy
Matteo G. Della Porta
IRCCS Azienda Ospedaliero - Universitaria di Bologna, Bologna, Italy
Enrico Giampieri & Gastone Castellani
Authors
- Francesco Casadei
- Patricia A. Apellániz
- Luciana Carota
- Gianluca Asti
- Matteo G. Della Porta
- Juan Parras
- Enrico Giampieri
- 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/.
About this article
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
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