Single-Trial Discrimination Of Eeg Signals For Stroke Patients: A General Multi-Way Analysis

Ye Liu,Mingfen Li,Hao Zhang,Junhua Li,Jie Jia,Yi Wu,Jianting Cao,Liqing Zhang
DOI: https://doi.org/10.1109/EMBC.2013.6609973
2013-01-01
Abstract:It has been demonstrated that Brain-Computer Interface (BCI), combined with Functional Electrical Stimulation (FES), is an effective and efficient way for post-stroke patients to restore motor function. However, traditional feature extraction methods, such as Common Spatial Pattern (CSP), do not work well for post-stroke patients' EEG data due to its irregular patterns. In this study, we introduce a novel tensor-based feature extraction algorithm, which takes both spatial-spectral-temporal features of EEG data into consideration. EEG data recorded from post-stroke patients is used for simulation experiments to assess the effectiveness of the proposed algorithm. The results show that the the proposed algorithm outperforms some traditional algorithms.
What problem does this paper attempt to address?