A Domain-Informed Multi-Objective Framework for EEG Channel Selection in Motor Imagery BCIs

2026-05-28

AI Summary

This study addresses the limitations of conventional EEG channel selection methods, which typically rely on single-objective criteria and struggle to simultaneously account for spatial correlation and task discriminability, often leading to suboptimal solutions. To overcome this, the authors propose a novel multi-objective optimization framework that integrates domain knowledge by jointly modeling spatial correlation—defined via a Gaussian kernel—and functional discriminability based on trial-wise task-related desynchronization. Pareto-optimal channel subsets are identified using evolutionary algorithms, including NSGA-II, MOPSO, and MOEA/D. Evaluated on the PhysioNet, OpenBMI, HighGamma, and BCI Competition IV-2a datasets, the approach achieves classification accuracies of 87%, 71%, 75%, and 65%, respectively, significantly outperforming existing single-objective and fixed-channel methods. The selected channels exhibit both compactness and physiological interpretability while reducing system complexity.

Abstract

Motor imagery (MI) classification using electroencephalography (EEG) signals is essential for advancing brain-computer interfaces (BCIs). Traditional EEG channel selection methods often face limitations, such as dependency on single-objective criteria and susceptibility to local optima. To address these challenges, this work proposes a multi-objective optimisation framework that employs non-dominated sorting genetic algorithm, multiple-objective particle swarm optimisation, and a multi-objective evolutionary algorithm based on decomposition. Our approach effectively balances spatial relevance, using a Gaussian kernel, and functional discriminability, which assesses intratrial task-related desynchronisation, thereby improving performance. We evaluated this framework on four EEG datasets: Physionet, OpenBMI, HighGamma, and BCIIV-2A. The proposed approach successfully identifies compact, relevant channel subsets concentrated around sensorimotor cortex regions linked to MI activity, addressing the prevalent challenges of dimensionality and complexity inherent to traditional techniques. Furthermore, the framework achieved classification performance of 87%, 71%, 75%, and 65% on the Physionet, OpenBMI, HighGamma, and BCIIV-2A datasets, respectively. By outperforming existing single-objective and accuracy-based methods, and those relying on fixed subsets, these findings demonstrate that this new multi-objective optimisation framework can enhance MI-based BCI performance while facilitating compact channel configurations with reduced computational complexity, making them better suited for wearable, portable, and real-time BCI applications.

Problem

• EEG channel selection
• motor imagery
• brain-computer interface
• multi-objective optimization
• dimensionality reduction

Innovation

• multi-objective optimization
• EEG channel selection
• motor imagery BCI
• functional discriminability
• spatial relevance