Abstract:Motor imagery, one of the main brain-computer interface (BCI) paradigms, has been extensively utilized in numerous BCI applications, such as the interaction between disabled people and external devices. Precise decoding, one of the most significant aspects of realizing efficient and stable interaction, has received a great deal of intensive research. However, the current decoding methods based on deep learning are still dominated by single-scale serial convolution, which leads to insufficient extraction of abundant information from motor imagery signals. To overcome such challenges, we propose a new end-to-end convolutional neural network based on multiscale spatial-temporal feature fusion (MSTFNet) for EEG classification of motor imagery. The architecture of MSTFNet consists of four distinct modules: feature enhancement module, multiscale temporal feature extraction module, spatial feature extraction module and feature fusion module, with the latter being further divided into the depthwise separable convolution block and efficient channel attention block. Moreover, we implement a straightforward yet potent data augmentation strategy to bolster the performance of MSTFNet significantly. To validate the performance of MSTFNet, we conduct cross-session experiments and leave-one-subject-out experiments. The cross-session experiment is conducted across two public datasets and one laboratory dataset. On the public datasets of BCI Competition IV 2a and BCI Competition IV 2b, MSTFNet achieves classification accuracies of 83.62% and 89.26%, respectively. On the laboratory dataset, MSTFNet achieves 86.68% classification accuracy. Besides, the leave-one-subject-out experiment is performed on the BCI Competition IV 2a dataset, and MSTFNet achieves 66.31% classification accuracy. These experimental results outperform several state-of-the-art methodologies, indicate the proposed MSTFNet's robust capability in decoding EEG signals associated with motor imagery.

T3SFNet: A Tuned Topological Temporal-Spatial Fusion Network for Motor Imagery with Rehabilitation Exoskeleton.

Motor Imagery Classification Based on Deep Convolutional Neural Network and Its Application in Exoskeleton Controlled by EEG

Multiscale Spatial-Temporal Feature Fusion Neural Network for Motor Imagery Brain-Computer Interfaces

EISATC-Fusion: Inception Self-Attention Temporal Convolutional Network Fusion for Motor Imagery EEG Decoding

TMSA-Net:A novel attention mechanism for improved motor imagery EEG signal processing

A Spatio-Temporal Interactive Attention Network for Motor Imagery EEG Decoding

EEG Decoding Method Based on Multi-Feature Information Fusion for Spinal Cord Injury

Corrupted EEG Self-Robust Motion Intention Recognition Method with Multi-Graph Fusion Enhanced Convolutional Network*

MI-MBFT: Superior Motor Imagery Decoding of Raw EEG Data Based on a Multi-Branch and Fusion Transformer Framework

A coarse-to-fine adaptive spatial–temporal graph convolution network with residuals for motor imagery decoding from the same limb

A Spatial-Spectral and Temporal Dual Prototype Network for Motor Imagery Brain-Computer Interface

EEG-Based Multi-Frequency Multilayer Network for Exploring the Brain State Evolution Underlying Motor Imagery

MCFHNet: Multi-Channel Fusion Hybrid Network for Efficient EEG-fNIRS Multi-modal Motor Imagery Decoding

Temporal-frequency-phase feature classification using 3D-convolutional neural networks for motor imagery and movement

MSATNet: multi-scale adaptive transformer network for motor imagery classification

Optimized Motor Imagery Paradigm Via Multimodal Stimulation and Explainable LSTM Model in Fnirs-Based BCI

Self-attention-based Convolutional Neural Network AND Time-frequency Common Spatial Pattern for enhanced Motor Imagery Classification

Spatial-Temporal Mamba Network for EEG-based Motor Imagery Classification

A Study of Unilateral Upper Limb Fine Motor Imagery Decoding Using Frequency-Band Attention Network

A Mutli-Scale Spatial-Temporal Convolutional Neural Network with Contrastive Learning for Motor Imagery EEG Classification

Spatiotemporal Feature Extraction Method Combining Common Spatial Pattern and Hybrid Spatial‐temporal Attention‐net in Motor Imagery–brain‐computer Interface System