Temporal-spatial convolutional residual network for decoding attempted movement related EEG signals of subjects with spinal cord injury

• Published in: Computers in biology and medicine Vol. 164; p. 107159
• Main Authors: Mirzabagherian, Hamed, Menhaj, Mohammad Bagher, Suratgar, Amir Abolfazl, Talebi, Nasibeh, Abbasi Sardari, Mohammad Reza, Sajedin, Atena
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.09.2023; Elsevier Limited
• Subjects: Algorithms; Artificial neural networks; Bayesian analysis; Brain; Brain computer interface; Brain injury; Classification; Computer applications; Convolutional neural network; Datasets; Deep learning; EEG; Electroencephalography; Electroencephalography (EEG); Feature extraction; Frequency ranges; Human-computer interface; Implants; Internal Medicine; Iterative methods; Mathematical models; Neural networks; Optimization; Other; Signal classification; Spinal cord injuries; Spinal cord injury; Deep learning; Brain computer interface; Electroencephalography (EEG); Spinal cord injury; Convolutional neural network
• ISSN: 0010-4825; 1879-0534; 1879-0534
• DOI: 10.1016/j.compbiomed.2023.107159

Brain Computer Interface (BCI) offers a promising approach to restoring hand functionality for people with cervical spinal cord injury (SCI). A reliable classification of brain activities based on appropriate flexibility in feature extraction could enhance BCI systems performance. In the present study, based on convolutional layers with temporal-spatial, Separable and Depthwise structures, we develop Temporal-Spatial Convolutional Residual Network)TSCR-Net(and Temporal-Spatial Convolutional Iterative Residual Network)TSCIR-Net(structures to classify electroencephalogram (EEG) signals. Using EEG signals in five different hand movement classes of SCI people, we compare the effectiveness of TSCIR-Net and TSCR-Net models with some competitive methods. We use the bayesian hyperparameter optimization algorithm to tune the hyperparameters of compact convolutional neural networks. In order to show the high generalizability of the proposed models, we compare the results of the models in different frequency ranges. Our proposed models decoded distinctive characteristics of different movement efforts and obtained higher classification accuracy than previous deep neural networks. Our findings indicate that TSCIR-Net and TSCR-Net models fulfills a better classification accuracy of 71.11%, and 64.55% for EEG_All and 57.74%, and 67.87% for EEG_Low frequency data sets than the compared methods in the literature. •Propose compact deep convolutional networks for classifying challenging EEG data.•Classification of attempted movement EEG data of subjects with spinal cord injury.•Designing TSCRNet and TSCIRNet to reach better classification results by using designed Residual block.•Comparing other models to classify challenging SCI data due to evaluate the results.