README
This dataset contains EEG recordings from four subjects performing motor imagery tasks (left hand, right hand, and feet), originally published by Zhou et al. (2016). The data was reformatted into BIDS from its Zenodo version (https://zenodo.org/records/16534752), which was itself generated by MOABB (Mother of All BCI Benchmarks, https://github.com/NeuroTechX/moabb). The original study investigated a fully automated trial selection method for optimization of motor imagery based brain-computer interfaces.
Four participants each completed three recording sessions separated by days to months. Each session contained two consecutive runs with inter-run breaks. Each run comprised 75 trials (25 per class: left hand, right hand, and feet imagery), for a total of 450 trials per subject across all sessions. Trials began with an auditory cue, followed by a 5-second visual arrow stimulus indicating the motor imagery task to perform, then a 4-second rest period. EEG was recorded from 14 channels placed according to the extended 10/20 system (Fp1, Fp2, FC3, FCz, FC4, C3, Cz, C4, CP3, CPz, CP4, O1, Oz, O2) at a sampling frequency of 250 Hz with a 50 Hz power line frequency.
- 4 subjects (sub-1 through sub-4)
- 3 sessions per subject (ses-0, ses-1, ses-2)
- 2 runs per session (run-0, run-1)
- 24 EEG recordings total in EDF format
- 14 EEG channels, 250 Hz sampling rate
- 3 event types: left_hand (value=2), right_hand (value=3), feet (value=1)
- Electrode positions in CapTrak coordinate system
The data distributed here has undergone minimal preprocessing by MOABB prior to BIDS conversion:
- Extraction of the 14 EEG channels from the original recordings
- Annotation of motor imagery events (left_hand, right_hand, feet) with 5-second durations
- Resampling to 250 Hz
- Export to EDF format
This dataset was reformatted into BIDS from the Zenodo archive at https://zenodo.org/records/16534752. That archive was generated by MOABB v1.2.0 from the original data accompanying the publication. The original study and data are described in:
Zhou B, Wu X, Lv Z, Zhang L, Guo X (2016). A Fully Automated Trial Selection Method for Optimization of Motor Imagery Based Brain-Computer Interface. PLoS ONE 11(9): e0162657. https://doi.org/10.1371/journal.pone.0162657
Zhou B, Wu X, Lv Z, Zhang L, Guo X (2016). A Fully Automated Trial Selection Method for Optimization of Motor Imagery Based Brain-Computer Interface. PLoS ONE 11(9): e0162657. https://doi.org/10.1371/journal.pone.0162657
Appelhoff S, Sanderson M, Brooks T, et al. (2019). MNE-BIDS: Organizing electrophysiological data into the BIDS format and facilitating their analysis. Journal of Open Source Software 4: 1896. https://doi.org/10.21105/joss.01896
Pernet CR, Appelhoff S, Gorgolewski KJ, et al. (2019). EEG-BIDS, an extension to the brain imaging data structure for electroencephalography. Scientific Data, 6, 103. https://doi.org/10.1038/s41597-019-0104-8
Data curator for NEMAR version: Arnaud Delorme (UCSD, La Jolla, CA, USA)