Federated deep learning model for epilepsy seizure detection using electroencephalogram (EEG) signal

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: Accuracy: 98.91%; Precision: 98.82%; 'Security': 98.93% — all point estimates without variability measures or comparison effect sizes

• Primary Outcome: Seizure detection accuracy on three unnamed benchmark EEG datasets

• Nnt Or Sensitivity: Sensitivity and specificity not reported. False alarm rate not reported. No NNT or diagnostic accuracy metrics (AUROC, likelihood ratios) provided. These omissions critically limit clinical interpretability of the model's performance.

• Confidence Interval: Not reported

Clinical Application

Clinical feasibility is not assessed. Computational requirements for federated learning with homomorphic encryption are substantial and not evaluated in the context of hospital IT infrastructure. Real-time detection latency, integration with clinical EEG systems, and workflow compatibility are not addressed. The model has not been tested outside a simulation environment. This study has no direct applicability to current Australian clinical practice. The TGA has not evaluated or approved this system. It does not align with RACGP or Epilepsy Society of Australia guidelines for seizure management, which require validated diagnostic tools. The PBS does not fund computational EEG analysis tools of this nature. Australian neurophysiology laboratories operate under NATA accreditation standards that would require prospective clinical validation before any such system could be considered for clinical use. The federated learning privacy framework, while conceptually relevant to Australian Privacy Act requirements and the My Health Record system, has not been evaluated against these specific regulatory standards. Not determinable. The study does not describe the patient population from which EEG data were derived, including seizure types (focal, generalised, absence), age groups, medication status, or EEG recording parameters. Clinical applicability to any specific epilepsy population cannot be established.

Abstract

OBJECTIVES: Epilepsy is a chronic neurological disorder characterized by recurrent seizures due to abnormal brain activity, which affects individuals' health and quality of life. Traditional seizure detection methods face challenges related to data privacy and security as well as difficulty in fully capturing both temporal and spatial relationships within the Electroencephalography signal. To address these limitations, a Federated Learning Enabled Unified Transformer model is proposed. METHODS: The Federated Learning with Paillier Homomorphic Encryption is deployed for preserving data privacy and enabling collaborative model training. Adaptive noise filtering and Independent Component Analysis are deployed to remove the noise and artifacts. The Multi-Scale Wavelet Coefficient is applied for signal decomposition that effectively extracts seizure-related features by decomposing Electroencephalography signals into multiple sub-bands. The Hybrid Graph-Based Attention Framework integrated Edge-Enhanced Graph Convolutional Networks for spatial feature extraction and Spectral Graph Attention for frequency-based feature selection, and refining feature representation for improving the classification accuracy. Further, the proposed technique utilizes a Unified Transformer model for seizure classification that efficiently captures temporal and spatial dependencies in Electroencephalography signals. RESULTS: The proposed model is validated on three datasets and attains 98.91% accuracy, 98.93% security, and 98.82% precision. The simulation outcomes indicate that the Federated Learning Enabled Unified Transformer model achieved outstanding performance when compared to existing models. DISCUSSION: The Federated Learning Enabled Unified Transformer model provided a superior performance by integrating Federated learning and hybrid Deep Learning models. It ensures that the model is more suitable for healthcare users.

References

1. 1.Abijith, G. R., Jothi, S. A., & Chandrasekar. (2025). Federated deep learning model for epilepsy seizure detection using electroencephalogram (EEG) signal. Neurological Research. Advance online publication. https://doi.org/10.1080/01616412.2025.2555516