Sliding Window Along With EEGNet-Based Prediction of EEG Motor Imagery

The need for repeated calibration and accounting for intersubject variability is a major challenge for the practical applications of a brain–computer interface (BCI). The problem becomes more challenging while decoding the brain signals of stroke patients due to altered neurodynamics caused by lesions. Recently, several deep learning architectures came into the picture although they often failed to produce superior accuracy compared to the traditional approaches and mostly do not follow an end-to-end architecture as they depend on custom features. However, a few of them have the promising ability to create more generalizable features in an end-to-end fashion such as the popular EEGNet architecture. Although EEGNet was applied for decoding stroke patients’ motor imagery (MI) data, its performance was as good as the traditional methods. In this study, we have augmented the EEGNet-based decoding by introducing a postprocessing step called the longest consecutive repetition (LCR) in a sliding window-based approach and named it EEGNet + LCR. The proposed approach was tested on a dataset of ten hemiparetic stroke patients’ MI dataset yielding superior performance against the only EEGNet and a more traditional approach such as common spatial pattern (CSP) + support vector machine (SVM) for both within- and cross-subject decoding of MI signals. We also observed comparable and satisfactory performance of the EEGNet + LCR in both the within- and cross-subject categories that are rarely found in the literature making it a promising candidate to realize practically feasible BCI for stroke rehabilitation.