EEG datasets for motor imagery brain–computer interface

Abstract

Background:Most investigators of brain–computer interface (BCI) research believe that BCI can be achieved through induced neuronal activity from the cortex, but not by evoked neuronal activity. Motor imagery (MI)–based BCI is one of the standard concepts of BCI, in that the user can generate induced activity by imagining motor movements. However, variations in performance over sessions and subjects are too severe to overcome easily; therefore, a basic understanding and investigation of BCI performance variation is necessary to find critical evidence of performance variation. Here we present not only EEG datasets for MI BCI from 52 subjects, but also the results of a psychological and physiological questionnaire, EMG datasets, the locations of 3D EEG electrodes, and EEGs for non-task-related states.Findings:We validated our EEG datasets by using the percentage of bad trials, event-related desynchronization/synchronization (ERD/ERS) analysis, and classification analysis. After conventional rejection of bad trials, we showed contralateral ERD and ipsilateral ERS in the somatosensory area, which are well-known patterns of MI. Finally, we showed that 73.08% of datasets (38 subjects) included reasonably discriminative information.Conclusions:Our EEG datasets included the information necessary to determine statistical significance; they consisted of well-discriminated datasets (38 subjects) and less-discriminative datasets. These may provide researchers with opportunities to investigate human factors related to MI BCI performance variation, and may also achieve subject-to-subject transfer by using metadata, including a questionnaire, EEG coordinates, and EEGs for non-task-related states.

Deep Neural Network with Joint Distribution Matching for Cross-Subject Motor Imagery Brain-Computer Interfaces

“…They vary a lot across subjects and show very bad separation among subjects. The second dataset adopted in this paper was supplied by Handong Global University [47], called GigaDataset (http://gigadb.org/dataset/100295) in our paper. 52 healthy subjects (26 males, 26 females; mean age: 24:8 ± 3:86 years, called subj1~subj52) participated.…”

“…Outputs from the fifth layer (flatten layer) are taken as initial features. After that, kernel principal component analysis (KPCA) is adopted and the dimension is reduced to 2 [47,48]. The discriminative lines are obtained by linear discriminative analysis on the two-dimensional data.…”

“…GrazA [46] can be found at URL: http://www.bbci.de/competition/iv/. The GigaDataset [47] in this paper can be found at URL: http://gigadb.org/ dataset/100295.…”

Low Complexity Automatic Stationary Wavelet Transform for Elimination of Eye Blinks from EEG

“…EEG-MAT database had already been filtered and the data was clean. To generate simulated EEG signals contaminated with eye blinks, 54 real eyeblink signals from the BCI experiment for motor imagery movement of the left and right hand [47], and the BCI Competition 2008 (Graz Data Sets 2a) [48] have been used. The eye blink signals were band-pass filtered between 0.1 and 3 Hz and resampled to 256 and 500 Hz.…”

“…In order to assess the performance of the proposed method on real EEG signals contaminated with eye blinks, 8 EEG signals from the BCI experiment for motor imagery movement of the left and right hand ( f s = 512 Hz) [47] and 8 EEG signals from the BCI Competition 2008 (Graz Data Sets 2a) ( f s = 250 Hz) [48] with length of 60 s have been used. Each signal was divided into 6 windows of 10 s and then the denoising methods were applied.…”