Addressing data heterogeneity in distributed medical imaging with heterosync learning

Hang-Tong Hu, Ming- De Li, Xin-Xin Lin, Meng-Yao Cai, Shuai Liu, Shao-Hong Wu, Wen-Juan Tong, Feng-Yu Ye, Jin-Bo Hu, Wei-Ping Ke, Li-Da Chen, Hong Yang, Guang-Jian Liu, Hai-Bo Wang, Ming- De Lu, Qing-Hua Huang (), Ming Kuang () and Wei Wang ()
Additional contact information
Hang-Tong Hu: the First Affiliated Hospital of Sun Yat-sen University
Ming- De Li: the First Affiliated Hospital of Sun Yat-sen University
Xin-Xin Lin: the First Affiliated Hospital of Sun Yat-sen University
Meng-Yao Cai: the First Affiliated Hospital of Sun Yat-sen University
Shuai Liu: Guangxi Minzu University
Shao-Hong Wu: the First Affiliated Hospital of Sun Yat-sen University
Wen-Juan Tong: the First Affiliated Hospital of Sun Yat-sen University
Feng-Yu Ye: Guangxi Minzu University
Jin-Bo Hu: Guangxi Minzu University
Wei-Ping Ke: the First Affiliated Hospital of Sun Yat-sen University
Li-Da Chen: the First Affiliated Hospital of Sun Yat-sen University
Hong Yang: the First Affiliated Hospital of Guangxi Medical University
Guang-Jian Liu: the Sixth Affiliated Hospital of Sun Yat-sen University (Guangdong Gastrointestinal Hospital)
Hai-Bo Wang: the First Affiliated Hospital of Sun Yat-Sen University
Ming- De Lu: the First Affiliated Hospital of Sun Yat-sen University
Qing-Hua Huang: Northwestern Polytechnical University
Ming Kuang: the First Affiliated Hospital of Sun Yat-sen University
Wei Wang: the First Affiliated Hospital of Sun Yat-sen University

Nature Communications, 2025, vol. 16, issue 1, 1-11

Abstract: Abstract Data heterogeneity critically limits distributed artificial intelligence (AI) in medical imaging. We propose HeteroSync Learning (HSL), a privacy-preserving framework that addresses heterogeneity through: (1) Shared Anchor Task (SAT) for cross-node representation alignment, and (2) an Auxiliary Learning Architecture coordinating SAT with local primary tasks. Validated via large-scale simulations (feature/label/quantity/combined heterogeneity) and a real-world multi-center thyroid cancer study, HSL outperforms local learning, 12 benchmark methods (FedAvg, FedProx, SplitAVG, FedRCL, FedCOME, etc.), and foundation models (e.g., CLIP) by better stability and up to 40% in area under the curve (AUC), matching central learning performance. HSL achieves 0.846 AUC on the out-of-distribution pediatric thyroid cancer data (outperforming others by 5.1-28.2%), demonstrating superior generalization. Visualizations confirm HSL successfully homogenizes heterogeneous distributions. This work provides an effective solution for distributed medical AI, enabling equitable collaboration across institutions and advancing healthcare AI democratization.

Date: 2025
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.nature.com/articles/s41467-025-64459-y Abstract (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:nat:natcom:v:16:y:2025:i:1:d:10.1038_s41467-025-64459-y

Ordering information: This journal article can be ordered from
https://www.nature.com/ncomms/

DOI: 10.1038/s41467-025-64459-y

Access Statistics for this article

Nature Communications is currently edited by Nathalie Le Bot, Enda Bergin and Fiona Gillespie

More articles in Nature Communications from Nature
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().