Hybrid CNN-GRU Models for Improved EEG Motor Imagery Classification DOI Creative Commons

Mouna Bouchane,

Wei Guo, Shuojin Yang

et al.

Sensors, Journal Year: 2025, Volume and Issue: 25(5), P. 1399 - 1399

Published: Feb. 25, 2025

Brain–computer interfaces (BCIs) based on electroencephalography (EEG) enable neural activity interpretation for device control, with motor imagery (MI) serving as a key paradigm decoding imagined movements. Efficient feature extraction from raw EEG signals is essential to improve classification accuracy while minimizing reliance extensive preprocessing. In this study, we introduce new hybrid architectures enhance MI using data augmentation and limited number of channels. The first model combines shallow convolutional network gated recurrent unit (CNN-GRU), the second incorporates bidirectional (CNN-Bi-GRU). Evaluated publicly available PhysioNet dataset, CNN-GRU classifier achieved peak mean rates 99.71%, 99.73%, 99.61%, 99.86% tasks involving left fist (LF), right (RF), both fists (LRF), feet (BF), respectively. experimental results provide compelling evidence that our proposed models outperform current state-of-the-art methods, underscoring their efficiency small-scale datasets. CNN-Bi-GRU exhibit superior predictive reliability, offering faster, cost-effective solution user-adaptable MI-BCI applications.

Language: Английский

Hybrid CNN-GRU Models for Improved EEG Motor Imagery Classification DOI Creative Commons

Mouna Bouchane,

Wei Guo, Shuojin Yang

et al.

Sensors, Journal Year: 2025, Volume and Issue: 25(5), P. 1399 - 1399

Published: Feb. 25, 2025

Brain–computer interfaces (BCIs) based on electroencephalography (EEG) enable neural activity interpretation for device control, with motor imagery (MI) serving as a key paradigm decoding imagined movements. Efficient feature extraction from raw EEG signals is essential to improve classification accuracy while minimizing reliance extensive preprocessing. In this study, we introduce new hybrid architectures enhance MI using data augmentation and limited number of channels. The first model combines shallow convolutional network gated recurrent unit (CNN-GRU), the second incorporates bidirectional (CNN-Bi-GRU). Evaluated publicly available PhysioNet dataset, CNN-GRU classifier achieved peak mean rates 99.71%, 99.73%, 99.61%, 99.86% tasks involving left fist (LF), right (RF), both fists (LRF), feet (BF), respectively. experimental results provide compelling evidence that our proposed models outperform current state-of-the-art methods, underscoring their efficiency small-scale datasets. CNN-Bi-GRU exhibit superior predictive reliability, offering faster, cost-effective solution user-adaptable MI-BCI applications.

Language: Английский

Citations

1