A Convolutional Neural Network Based Graph Theory for EEG Classification of Motor Imagery

Electroencephalography (EEG) is widely used due to their high temporal resolution. However, accurate classification remains a challenge due to the non-stationary nature of the EEG and the complexity of feature extraction. This study presents a convolutional neural network based graph theory to improve the classification accuracy of motor imagery (MI) EEG. Graph theory is used to construct spatial relationship models between EEG electrodes. Neural networks are used for classification, and residual connections are added to address node similarity issues that arise with increasing network complexity. The experimental results demonstrate a significant improvement in classification accuracy compared to the traditional method. Specifically, the accuracy of the test dataset exceeds 70.83% while also reducing computational complexity. These results demonstrate the robustness and generalize ability of the method. The proposed method achieves superior classification accuracy compared to existing techniques through graph theory algorithms and residual connections for EEG-based BCI. This study presents a method for improving the classification accuracy of MI tasks using EEG. The combination of graph theory algorithms and residual connections provides a promising avenue for future research in this field, allowing for more accurate and reliable BCI technologies.