Tensor-CSPNet: a novel geometric deep learning framework for motor imagery classification
Deep learning (DL) has been widely investigated in a vast majority of applications in electroencephalography (EEG)-based brain-computer interfaces (BCIs), especially for motor imagery (MI) classification in the past five years. The mainstream DL methodology for the MI-EEG classification exploits the...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2023
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/164529 |
| _version_ | 1811693294960246784 |
|---|---|
| author | Ju, Ce Guan, Cuntai |
| author2 | School of Computer Science and Engineering |
| author_facet | School of Computer Science and Engineering Ju, Ce Guan, Cuntai |
| author_sort | Ju, Ce |
| collection | NTU |
| description | Deep learning (DL) has been widely investigated in a vast majority of applications in electroencephalography (EEG)-based brain-computer interfaces (BCIs), especially for motor imagery (MI) classification in the past five years. The mainstream DL methodology for the MI-EEG classification exploits the temporospatial patterns of EEG signals using convolutional neural networks (CNNs), which have been particularly successful in visual images. However, since the statistical characteristics of visual images depart radically from EEG signals, a natural question arises whether an alternative network architecture exists apart from CNNs. To address this question, we propose a novel geometric DL (GDL) framework called Tensor-CSPNet, which characterizes spatial covariance matrices derived from EEG signals on symmetric positive definite (SPD) manifolds and fully captures the temporospatiofrequency patterns using existing deep neural networks on SPD manifolds, integrating with experiences from many successful MI-EEG classifiers to optimize the framework. In the experiments, Tensor-CSPNet attains or slightly outperforms the current state-of-the-art performance on the cross-validation and holdout scenarios in two commonly used MI-EEG datasets. Moreover, the visualization and interpretability analyses also exhibit the validity of Tensor-CSPNet for the MI-EEG classification. To conclude, in this study, we provide a feasible answer to the question by generalizing the DL methodologies on SPD manifolds, which indicates the start of a specific GDL methodology for the MI-EEG classification. |
| first_indexed | 2024-10-01T06:49:24Z |
| format | Journal Article |
| id | ntu-10356/164529 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T06:49:24Z |
| publishDate | 2023 |
| record_format | dspace |
| spelling | ntu-10356/1645292023-01-31T02:08:09Z Tensor-CSPNet: a novel geometric deep learning framework for motor imagery classification Ju, Ce Guan, Cuntai School of Computer Science and Engineering S-Lab Engineering::Computer science and engineering Geometric Deep Learning Motor Imagery Classification Deep learning (DL) has been widely investigated in a vast majority of applications in electroencephalography (EEG)-based brain-computer interfaces (BCIs), especially for motor imagery (MI) classification in the past five years. The mainstream DL methodology for the MI-EEG classification exploits the temporospatial patterns of EEG signals using convolutional neural networks (CNNs), which have been particularly successful in visual images. However, since the statistical characteristics of visual images depart radically from EEG signals, a natural question arises whether an alternative network architecture exists apart from CNNs. To address this question, we propose a novel geometric DL (GDL) framework called Tensor-CSPNet, which characterizes spatial covariance matrices derived from EEG signals on symmetric positive definite (SPD) manifolds and fully captures the temporospatiofrequency patterns using existing deep neural networks on SPD manifolds, integrating with experiences from many successful MI-EEG classifiers to optimize the framework. In the experiments, Tensor-CSPNet attains or slightly outperforms the current state-of-the-art performance on the cross-validation and holdout scenarios in two commonly used MI-EEG datasets. Moreover, the visualization and interpretability analyses also exhibit the validity of Tensor-CSPNet for the MI-EEG classification. To conclude, in this study, we provide a feasible answer to the question by generalizing the DL methodologies on SPD manifolds, which indicates the start of a specific GDL methodology for the MI-EEG classification. Published version This work was supported in part by the RIE2020 Industry Alignment Fund–Industry Collaboration Projects (IAF-ICP) Funding Initiative and in part by the RIE2020 AME Programmatic Fund, Singapore, under Grant A20G8b0102. 2023-01-31T02:08:09Z 2023-01-31T02:08:09Z 2022 Journal Article Ju, C. & Guan, C. (2022). Tensor-CSPNet: a novel geometric deep learning framework for motor imagery classification. IEEE Transactions On Neural Networks and Learning Systems, 1-15. https://dx.doi.org/10.1109/TNNLS.2022.3172108 2162-237X https://hdl.handle.net/10356/164529 10.1109/TNNLS.2022.3172108 35749326 2-s2.0-85133748336 1 15 en A20G8b0102 IEEE Transactions on Neural Networks and Learning Systems © 2022 The Authors. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/. application/pdf |
| spellingShingle | Engineering::Computer science and engineering Geometric Deep Learning Motor Imagery Classification Ju, Ce Guan, Cuntai Tensor-CSPNet: a novel geometric deep learning framework for motor imagery classification |
| title | Tensor-CSPNet: a novel geometric deep learning framework for motor imagery classification |
| title_full | Tensor-CSPNet: a novel geometric deep learning framework for motor imagery classification |
| title_fullStr | Tensor-CSPNet: a novel geometric deep learning framework for motor imagery classification |
| title_full_unstemmed | Tensor-CSPNet: a novel geometric deep learning framework for motor imagery classification |
| title_short | Tensor-CSPNet: a novel geometric deep learning framework for motor imagery classification |
| title_sort | tensor cspnet a novel geometric deep learning framework for motor imagery classification |
| topic | Engineering::Computer science and engineering Geometric Deep Learning Motor Imagery Classification |
| url | https://hdl.handle.net/10356/164529 |
| work_keys_str_mv | AT juce tensorcspnetanovelgeometricdeeplearningframeworkformotorimageryclassification AT guancuntai tensorcspnetanovelgeometricdeeplearningframeworkformotorimageryclassification |