Hilbert transform-based event-related patterns for motor imagery brain computer interface

Event-related patterns (EPs) play an essential role in detecting motor imagery (MI) movements of the subject. Due to the difference in the spatial and temporal distribution of brain signals among the subjects, the extraction of EP is a major issue. To rectify this problem, the Hilbert transform (HT) was used for the detection of EPs, and the machine learning (ML) models were implemented for decoding MI movements. The proposed method comprises two steps: initially, μ (8–12 Hz) and β (12–30 Hz) frequency bands were extracted from the raw electroencephalogram (EEG) signal. The HT was implemented on extracted μ and β bands signals and the EPs were calculated. Finally, the EPs were fed into two ML models such as support vector machine (SVM) and logistic regression (LR) for the detection of MI movements. The proposed method was tested on two benchmark datasets (BCI competition-III and IV). The results show that the mean classification accuracy (%CA) and Cohen's kappa coefficient (K) for BCI competition-III and IV were 86.11% & 0.72 and 82.50% & 0.65 respectively, which are higher than several existing methods.