科研速览 · Science Skim继续刷下去 · Keep skimming →
◆ Neuroscience2026-08-04

Synergistic EEG signal processing for brain-computer interfaces using hybrid MothCray optimization and deep learning.

Annu Kumari, Damodar Reddy Edla, Dharavath Ramesh, Vaishali R Shirodkar

原始摘要(英文原文)· Original abstract
Electroencephalography (EEG) based brain-computer interface (BCI) systems require optimal channel selection to achieve high signal quality, reduced setup complexity, and robust usability. This study introduces a comprehensive signal-processing framework to improve the efficiency and accuracy of EEG-driven BCIs. Signal pre-processing incorporates standard EEG denoising techniques, including notch filtering, independent component analysis (ICA), and segmentation into temporal windows to mitigate artifacts and enhance data quality. The channel selection phase employs the novel mothcray optimization (MCO) algorithm, a hybrid approach that integrates moth flame optimization and crayfish optimization to identify the most informative channels. Feature extraction encompasses time-domain statistics, frequency-domain attributes, and connectivity measures, enabling a rich representation of underlying neural dynamics. Classification is performed using an advanced deep neural network architecture tailored for the spatial-temporal characteristics of EEG data. Validated on the BCI competition IV dataset IIa, the proposed model achieves an accuracy rate of 93.92%, outperforming established methods and underscoring its effectiveness. The proposed MCO-based channel selection, multi-domain feature fusion, and enhanced EEGNet classifier demonstrate a significant step forward for practical BCI deployment that requires fast setup, robust operation, and efficient use of resources. It is especially advantageous in applications that rely on a reduced number of EEG channels to enable more comfortable and portable headsets, must sustain stable performance under varying recording conditions in home- or clinic-based neurorehabilitation and assistive communication settings.
读原文 · Read the paper ↗

AI 追问PRO

登录后使用 AI 追问

讨论区

登录后参与讨论

相关论文 · Related

Synergistic EEG signal processing for brain-computer interfaces using hybrid MothCray optimization and deep learning. — 科研速览 Science Skim