We used an Ultracortex Mark IV headset (OpenBCI, New York, NY, USA). Eight electrodes (FP1, FP2, C3, C4, P7, P8, O1, and O2) and reference electrodes on both ear lobes were applied according to the 10–20 international system for electrode placement. OpenBCI has been used previously to predict the amplitude modulation of steady-state, visually evoked potentials signals with a single electrode [45]. OpenBCI has been used to develop EEG-based applications, including a device for disabled people [46], and to assess P3, N2, and FRN components for performance monitoring [47]. In a comparison of dry and wet electrode EEG systems, the dry electrode device was found to be more robust to 50 Hz line noise, less sensitive to electromagnetic interference, and useful for self-application and home usage [48]. The 250 Hz sampling frequency was used to acquire the EEG. Pre-processing was carried out to remove the noise and artefacts from the data. We used EEGLAB toolbox for preprocessing and FastICA algorithm was used to remove the artifacts from the electrooculography (EOG) and electromyography (EMG) artifacts out of the EEG signals [49].
