The PFED protocol achieved 87.77% accuracy in epilepsy detection with 5 clients, closely matching the 88.75% accuracy of standard federated learning while providing strong privacy protection.
The proposed PFED protocol enables secure, privacy-preserving federated learning for EEG-based epilepsy detection without compromising model accuracy.
Absolute Event Rate: 87.77% vs 88.75%
The convergence of Generative Artificial Intelligence and biometric recognition is reshaping modern healthcare. It enables more adaptive and intelligent human–machine interactions. Epilepsy, a common neurological disorder affecting millions worldwide, relies heavily on electroencephalography (EEG) signals for diagnosis and monitoring. Wearable consumer devices with EEG sensors support continuous physiological data collection. However, transmitting sensitive biometric data to centralized servers introduces serious privacy and security risks. Federated learning (FL) provides a distributed training framework that keeps raw data on local devices. Despite this advantage, existing FL methods remain vulnerable to gradient leakage attacks, where adversaries may infer private biometric information from shared model updates. To address this issue, we propose PFED, a privacy-preserving federated learning protocol that combines randomized group interaction with volunteer-assisted secure aggregation. The proposed method effectively obfuscates gradient information while maintaining model utility. Experiments on the public CHB-MIT EEG dataset show that PFED achieves reliable epilepsy detection performance and strong privacy protection. These results demonstrate its potential for secure AI-enabled biometric healthcare applications in consumer device environments.
Xu et al. (Thu,) conducted a other in Epilepsy (n=23). PFED (Privacy-Preserving Federated Learning protocol) vs. Standard Federated Learning (FED) was evaluated on Model accuracy for epilepsy detection (with 5 clients). The PFED protocol achieved 87.77% accuracy in epilepsy detection with 5 clients, closely matching the 88.75% accuracy of standard federated learning while providing strong privacy protection.