A fused multi-subfrequency bands and CBAM SSVEP-BCI classification method based on convolutional neural network
Abstract For the brain-computer interface (BCI) system based on steady-state visual evoked potential (SSVEP), it is difficult to obtain satisfactory classification performance for short-time window SSVEP signals by traditional methods. In this paper, a fused multi-subfrequency bands and convolutiona...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-04-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-024-59348-1 |
| _version_ | 1797209331173687296 |
|---|---|
| author | Dongyang Lei Chaoyi Dong Hongfei Guo Pengfei Ma Huanzi Liu Naqin Bao Hongzhuo Kang Xiaoyan Chen Yi Wu |
| author_facet | Dongyang Lei Chaoyi Dong Hongfei Guo Pengfei Ma Huanzi Liu Naqin Bao Hongzhuo Kang Xiaoyan Chen Yi Wu |
| author_sort | Dongyang Lei |
| collection | DOAJ |
| description | Abstract For the brain-computer interface (BCI) system based on steady-state visual evoked potential (SSVEP), it is difficult to obtain satisfactory classification performance for short-time window SSVEP signals by traditional methods. In this paper, a fused multi-subfrequency bands and convolutional block attention module (CBAM) classification method based on convolutional neural network (CBAM-CNN) is proposed for discerning SSVEP-BCI tasks. This method extracts multi-subfrequency bands SSVEP signals as the initial input of the network model, and then carries out feature fusion on all feature inputs. In addition, CBAM is embedded in both parts of the initial input and feature fusion for adaptive feature refinement. To verify the effectiveness of the proposed method, this study uses the datasets of Inner Mongolia University of Technology (IMUT) and Tsinghua University (THU) to evaluate the performance of the proposed method. The experimental results show that the highest accuracy of CBAM-CNN reaches 0.9813 percentage point (pp). Within 0.1–2 s time window, the accuracy of CBAM-CNN is 0.0201–0.5388 (pp) higher than that of CNN, CCA-CWT-SVM, CCA-SVM, CCA-GNB, FBCCA, and CCA. Especially in the short-time window range of 0.1–1 s, the performance advantage of CBAM-CNN is more significant. The maximum information transmission rate (ITR) of CBAM-CNN is 503.87 bit/min, which is 227.53 bit/min-503.41 bit/min higher than the above six EEG decoding methods. The study further results show that CBAM-CNN has potential application value in SSVEP decoding. |
| first_indexed | 2024-04-24T09:53:00Z |
| format | Article |
| id | doaj.art-2bf512d05af449a2ae531713a175087b |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-24T09:53:00Z |
| publishDate | 2024-04-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-2bf512d05af449a2ae531713a175087b2024-04-14T11:15:15ZengNature PortfolioScientific Reports2045-23222024-04-0114111110.1038/s41598-024-59348-1A fused multi-subfrequency bands and CBAM SSVEP-BCI classification method based on convolutional neural networkDongyang Lei0Chaoyi Dong1Hongfei Guo2Pengfei Ma3Huanzi Liu4Naqin Bao5Hongzhuo Kang6Xiaoyan Chen7Yi Wu8College of Electric Power, Inner Mongolia University of TechnologyCollege of Electric Power, Inner Mongolia University of TechnologyInner Mongolia Academy of Science and TechnologyCollege of Electric Power, Inner Mongolia University of TechnologyCollege of Electric Power, Inner Mongolia University of TechnologyCollege of Electric Power, Inner Mongolia University of TechnologyCollege of Electric Power, Inner Mongolia University of TechnologyCollege of Electric Power, Inner Mongolia University of TechnologyInner Mongolia Academy of Science and TechnologyAbstract For the brain-computer interface (BCI) system based on steady-state visual evoked potential (SSVEP), it is difficult to obtain satisfactory classification performance for short-time window SSVEP signals by traditional methods. In this paper, a fused multi-subfrequency bands and convolutional block attention module (CBAM) classification method based on convolutional neural network (CBAM-CNN) is proposed for discerning SSVEP-BCI tasks. This method extracts multi-subfrequency bands SSVEP signals as the initial input of the network model, and then carries out feature fusion on all feature inputs. In addition, CBAM is embedded in both parts of the initial input and feature fusion for adaptive feature refinement. To verify the effectiveness of the proposed method, this study uses the datasets of Inner Mongolia University of Technology (IMUT) and Tsinghua University (THU) to evaluate the performance of the proposed method. The experimental results show that the highest accuracy of CBAM-CNN reaches 0.9813 percentage point (pp). Within 0.1–2 s time window, the accuracy of CBAM-CNN is 0.0201–0.5388 (pp) higher than that of CNN, CCA-CWT-SVM, CCA-SVM, CCA-GNB, FBCCA, and CCA. Especially in the short-time window range of 0.1–1 s, the performance advantage of CBAM-CNN is more significant. The maximum information transmission rate (ITR) of CBAM-CNN is 503.87 bit/min, which is 227.53 bit/min-503.41 bit/min higher than the above six EEG decoding methods. The study further results show that CBAM-CNN has potential application value in SSVEP decoding.https://doi.org/10.1038/s41598-024-59348-1 |
| spellingShingle | Dongyang Lei Chaoyi Dong Hongfei Guo Pengfei Ma Huanzi Liu Naqin Bao Hongzhuo Kang Xiaoyan Chen Yi Wu A fused multi-subfrequency bands and CBAM SSVEP-BCI classification method based on convolutional neural network Scientific Reports |
| title | A fused multi-subfrequency bands and CBAM SSVEP-BCI classification method based on convolutional neural network |
| title_full | A fused multi-subfrequency bands and CBAM SSVEP-BCI classification method based on convolutional neural network |
| title_fullStr | A fused multi-subfrequency bands and CBAM SSVEP-BCI classification method based on convolutional neural network |
| title_full_unstemmed | A fused multi-subfrequency bands and CBAM SSVEP-BCI classification method based on convolutional neural network |
| title_short | A fused multi-subfrequency bands and CBAM SSVEP-BCI classification method based on convolutional neural network |
| title_sort | fused multi subfrequency bands and cbam ssvep bci classification method based on convolutional neural network |
| url | https://doi.org/10.1038/s41598-024-59348-1 |
| work_keys_str_mv | AT dongyanglei afusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT chaoyidong afusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT hongfeiguo afusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT pengfeima afusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT huanziliu afusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT naqinbao afusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT hongzhuokang afusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT xiaoyanchen afusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT yiwu afusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT dongyanglei fusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT chaoyidong fusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT hongfeiguo fusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT pengfeima fusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT huanziliu fusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT naqinbao fusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT hongzhuokang fusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT xiaoyanchen fusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork AT yiwu fusedmultisubfrequencybandsandcbamssvepbciclassificationmethodbasedonconvolutionalneuralnetwork |