Semi-Supervised Privacy-Preserving EEG-Based Motor Imagery Classification via Self and Adversarial Training

Electroencephalogram (EEG)-based motor imagery (MI) signals are frequently used in brain-computer interfaces (BCIs) due to their wide applications in the rehabilitation field. However, cross-subject variations often result in a model trained on one participant failing when applied to another. Additionally, privacy concerns regarding sensitive health and mental information in EEG-based MI signals further complicate the situation. Source-free domain adaptation aims to address these cross-subject variations by transferring knowledge from a source domain (i.e., a previous participant) to a target domain (i.e., a new participant) without accessing sensitive source data. However, source-free unsupervised domain adaptation models often face issues with incorrect pseudo-labels, which can lead to unstable and ineffective adaptation. To address this, we propose a source-free semi-supervised domain adaptation algorithm for EEG-based MI signal classification. This algorithm tackles noise accumulation caused by incorrect pseudo-labels while effectively handling data distribution variations and privacy concerns, similar to source-free unsupervised domain adaptation models. Specifically, we train the classifier head using only a limited amount of labeled target data to prevent noise accumulation, and generate pseudo-labels for the unlabeled target data. Furthermore, we introduce an independent self-training head that learns better representations using the generated pseudo-labels, mitigating overfitting caused by the limited labeled target data. Additionally, we design an adversarial head that plays a minimax game to extract more discriminative feature representations from the unlabeled target data. Extensive experiments on three benchmark datasets, compared with eighteen state-of-the-art SFDA methods, demonstrate the superiority of our approach. Note to Practitioners—The practical problem addressed in this work is the challenge of classifying electroencephalogram (EEG)-based motor imagery (MI) signals in brain-computer interfaces (BCIs) across different subjects, which is crucial for applications such as stroke and paralysis recovery. Our solution introduces a source-free semi-supervised domain adaptation algorithm that leverages a small amount of labeled target data to prevent noise accumulation caused by incorrect pseudo-labels, a common issue in unsupervised domain adaptation. By incorporating a self-training head and an adversarial head, we improve feature representation and reduce overfitting, leading to more accurate and reliable classification. This approach enhances the efficiency and reliability of BCI systems by reducing the need for extensive labeled data, ensuring privacy, and improving generalization capabilities across different subjects. The key insights include proposing the first semi-supervised source-free domain adaptation method for MI signal classification, developing a self-training head to mitigate noise accumulation, and designing an adversarial head to optimize feature extraction adversarially. Future work could explore further reducing computational cost, extending the method to other BCI signals, and investigating more advanced neural network architectures. The results can be extended to real-time BCI systems, personalized medical diagnostics, and enhancing AI-driven health monitoring systems.