Federated Learning for Epileptic Seizure Prediction Across Heterogeneous EEG Datasets

2025-08-11

AI Summary

In multi-center epilepsy seizure prediction, privacy compliance and non-IID data heterogeneity introduce model bias, particularly disadvantaging small-scale clinical sites. Method: We propose a randomized subgroup aggregation strategy for federated learning—replacing standard FedAvg—to ensure equitable client-level contribution during model aggregation and mitigate dataset-dominance bias. Our approach integrates single-channel EEG feature extraction, privacy-preserving global normalization, and a refined aggregation mechanism. Contribution/Results: Evaluated across four heterogeneous EEG datasets, the method achieves strong cross-center generalization, with macro-average accuracy of 77.1% and overall accuracy of 80.0%. Notably, performance improves significantly for small centers (e.g., Helsinki and NCH), demonstrating enhanced fairness, robustness, and privacy security without compromising predictive efficacy.

Technology Category

• Machine Learning: Privacy
• Computer Vision: Bias, Fairness & Privacy
• Multiagent Systems: Adversarial Agents

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Security and Privacy: Large-scale security measurements
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Developing accurate and generalizable epileptic seizure prediction models from electroencephalography (EEG) data across multiple clinical sites is hindered by patient privacy regulations and significant data heterogeneity (non-IID characteristics). Federated Learning (FL) offers a privacy-preserving framework for collaborative training, but standard aggregation methods like Federated Averaging (FedAvg) can be biased by dominant datasets in heterogeneous settings. This paper investigates FL for seizure prediction using a single EEG channel across four diverse public datasets (Siena, CHB-MIT, Helsinki, NCH), representing distinct patient populations (adult, pediatric, neonate) and recording conditions. We implement privacy-preserving global normalization and propose a Random Subset Aggregation strategy, where each client trains on a fixed-size random subset of its data per round, ensuring equal contribution during aggregation. Our results show that locally trained models fail to generalize across sites, and standard weighted FedAvg yields highly skewed performance (e.g., 89.0% accuracy on CHB-MIT but only 50.8% on Helsinki and 50.6% on NCH). In contrast, Random Subset Aggregation significantly improves performance on under-represented clients (accuracy increases to 81.7% on Helsinki and 68.7% on NCH) and achieves a superior macro-average accuracy of 77.1% and pooled accuracy of 80.0% across all sites, demonstrating a more robust and fair global model. This work highlights the potential of balanced FL approaches for building effective and generalizable seizure prediction systems in realistic, heterogeneous multi-hospital environments while respecting data privacy.

Problem

• Developing accurate epileptic seizure prediction models across diverse EEG datasets
• Addressing data heterogeneity and privacy in federated learning for EEG analysis
• Improving fairness and robustness in multi-site seizure prediction using FL

Innovation

• Federated Learning for privacy-preserving EEG analysis
• Random Subset Aggregation to balance client contributions
• Global normalization for heterogeneous dataset compatibility