Federated Learning for Epileptic Seizure Prediction Across Heterogeneous EEG Datasets

TL;DR

This paper explores federated learning for epileptic seizure prediction across diverse EEG datasets, proposing a balanced aggregation method to improve model fairness and generalization while preserving patient privacy.

Contribution

It introduces a Random Subset Aggregation strategy and privacy-preserving normalization to enhance federated learning performance on heterogeneous EEG data.

Findings

Standard federated averaging is biased by dominant datasets.

Random Subset Aggregation improves under-represented clients' accuracy.

The proposed method achieves a macro-average accuracy of 77.1%.

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.