Data Sets
Short overview of the two data sets used for validating the proposed methods.
| Healthy Data Set [1] | Stroke patients' Data Set [2] |
|---|---|
| four different MI tasks, namely the imagination of movement of the left hand, right hand, both feet, and tongue. | two different MI tasks namely left- versus right-hand movement visualized with an exoskeleton. |
| 9 subjects with two sessions, namely training and test sessions held on different days. Each session comprises 72 trials per class making a total of 288 trials. | 10 hemiparetic stroke patients with two training sessions and an online feedback session each consisting of 40 trials randomly distributed equally into left- and right-hand tasks. |
| All signals were bandpass filtered from 0.5 to 100 Hz (with a 50-Hz notch filter enabled). The recorded EEG signals were sampled at a frequency of 250 Hz. There were 22 EEG channels. | The data acquisition was done at a 500-Hz sampling rate and initially bandpass filtered between 0.1 and 100 Hz with a notch filter at 50 Hz to cancel the power line noise. There were 12 EEG channels. |
Two different data sets are used, because stroke patients’ data have more nonstationarity due to altered neurodynamics and lack of engagement during the task.
C. Brunner, R. Leeb, G. Müller-Putz, A. Schlögl and G. Pfurtscheller, BCI Competition 2008–Graz data set A. Institute for Knowledge Discovery (Laboratory of Brain-Computer Interfaces), Graz University of Technology, vol. 16, pp. 1-6, 2018, link ↩︎
A. Chowdhury, H. Raza, Y. K. Meena, A. Dutta and G. Prasad, "Online Covariate Shift Detection-Based Adaptive Brain–Computer Interface to Trigger Hand Exoskeleton Feedback for Neuro-Rehabilitation," in IEEE Transactions on Cognitive and Developmental Systems, vol. 10, no. 4, pp. 1070-1080, Dec. 2018, doi: 10.1109/TCDS.2017.2787040. ↩︎