aiqbench analysis
This paper introduces FedTGNN-SS, a federated semi-supervised framework designed to address label scarcity and privacy constraints in predicting gestational diabetes using electronic health records. By combining prototype-guided pseudo-labeling with adaptive graph refinement, the method effectively leverages unlabeled data across hospital silos while exchanging only class-level centroids to preserve patient privacy. The approach demonstrates significant performance gains over existing baselines under extreme data scarcity conditions, offering a viable pathway for deploying robust clinical AI models without compromising data sovereignty.
Federated Semi-Supervised Graph Neural Networks with Prototype-Guided Pseudo-Labeling for Privacy-Preserving Gestational Diabetes Mellitus Prediction
arXiv:2605.01810v2 Announce Type: replace Abstract: Gestational Diabetes Mellitus (GDM) is a high-prevalence pregnancy complication that requires accurate early risk stratification to reduce maternal and fetal morbidity. However, real-world clinical deployment of machine learning is hindered by two coupled constraints: (i) label scarcity, where a large fraction of electronic health records (EHR) lack confirmed diagnostic labels, and (ii) data privacy, which prevents sharing patient-level data across hospitals. This paper proposes FedTGNN-SS, a privacy-preserving federated semi-supervised framework for clinical tabular EHR. Each hospital builds a local k-nearest-neighbor patient similarity graph and trains a topology-adaptive GNN encoder. To robustly exploit unlabeled records, FedTGNN-SS combines (1) prototype-guided pseudo-labeling with neighborhood agreement, (2) adaptive graph refinement that periodically updates the k-NN graph using learned embeddings, (3) clinical-aware consistency augmentation applied only to continuous variables, and (4) privacy-safe prototype sharing that exchanges only class-level centroids. Across three diabetes-related datasets (GDM: N = 3,525; Pima: N = 768; Early Stage: N = 520) under 10\%-80\% missing labels per silo, FedTGNN-SS achieves 56 significant wins ($p < 0.05$) against 11 federated baselines and attains strong AUROC under extreme scarcity (Pima: 0.8037 at 80\% missing, Early Stage: 0.9634 at 80\% missing).
Abstract:Gestational Diabetes Mellitus (GDM) is a high-prevalence pregnancy complication that requires accurate early risk stratification to reduce maternal and fetal morbidity. However, real-world clinical deployment of machine learning is hindered by two coupled constraints: (i) label scarcity, where a large fraction of electronic health records (EHR) lack confirmed diagnostic labels, and (ii) data privacy, which prevents sharing patient-level data across hospitals. This paper proposes FedTGNN-SS, a privacy-preserving federated semi-supervised framework for clinical tabular EHR. Each hospital builds a local k-nearest-neighbor patient similarity graph and trains a topology-adaptive GNN encoder. To robustly exploit unlabeled records, FedTGNN-SS combines (1) prototype-guided pseudo-labeling with neighborhood agreement, (2) adaptive graph refinement that periodically updates the k-NN graph using learned embeddings, (3) clinical-aware consistency augmentation applied only to continuous variables, and (4) privacy-safe prototype sharing that exchanges only class-level centroids. Across three diabetes-related datasets (GDM: N = 3,525; Pima: N = 768; Early Stage: N = 520) under 10\%-80\% missing labels per silo, FedTGNN-SS achieves 56 significant wins ($p < 0.05$) against 11 federated baselines and attains strong AUROC under extreme scarcity (Pima: 0.8037 at 80\% missing, Early Stage: 0.9634 at 80\% missing).
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2605.01810 [cs.LG]
(or arXiv:2605.01810v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2605.01810 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 282 words
Related coverage
Graph Neural Networks for Semi-Supervised Image Classification with Multi-Feature Aggregation
arXiv · Jun 17, 2026
GLL: A Differentiable Graph Learning Layer for Neural Networks
arXiv · Oct 15, 2026
Timestamp-Aware Spatio-Temporal Graph Contrastive Learning for Network Intrusion Detection
arXiv · Jun 17, 2026
Rethinking One-Shot Federated Graph Learning: Training-Free Statistical Estimation
arXiv · Sep 9, 2026