moabb.datasets.BNCI2020_002#

class moabb.datasets.BNCI2020_002(subjects=None, sessions=None)[source]#

BNCI 2020-002 Attention Shift (Covert Spatial Attention) dataset.

Dataset summary

#Subj	18
#Chan	31
#Trials / class	varies NT / T
Trials length	16 s
Freq	250 Hz
#Sessions	1

Participants

Population: healthy
Age: 27 (range: 19-38) years

Equipment

Amplifier: BrainAmp DC Amplifier
Electrodes: Ag/AgCl electrodes
Montage: extended 10-20
Reference: right mastoid

Preprocessing

Data state: raw
Bandpass filter: 1-12.5 Hz
Steps: re-referenced to average of left and right mastoid, 4th order zero-phase IIR Butterworth bandpass filter (1.0-12.5 Hz), resampled to 50 Hz, epoched from stimulus onset to 750 ms after
Re-reference: average of left and right mastoid

Data Access

DOI: 10.3389/fnins.2020.591777
Data URL: http://bnci-horizon-2020.eu/database/data-sets
Repository: BNCI Horizon

Experimental Protocol

Paradigm: covert spatial attention
Task type: binary decision
Feedback: visual (yes/no text)
Stimulus: colored crosses (green + and red x)

Found an issue with this dataset?

If you encounter any problems with this dataset (missing files, incorrect metadata, loading errors, etc.), please let us know!

Report an Issue on GitHub

Dataset from [1].

Dataset Description

This dataset contains EEG recordings from 18 healthy subjects performing a covert spatial attention task for brain-computer interface (BCI) control. The paradigm decodes binary decisions based on the N2pc component - a neurological marker reflecting attention to visual targets in specific hemispheres.

Subjects were presented with colored stimuli (red and green crosses) in left and right visual hemifields simultaneously. By covertly shifting attention to one side (left or right), subjects could indicate “yes” or “no” responses without any overt movement, enabling gaze-independent communication.

Participants

18 healthy subjects (10 female)
Age range: 19-38 years (mean 27 years)
All right-handed
Normal or corrected-to-normal vision
Location: Otto-von-Guericke University Magdeburg, Germany

Recording Details

Equipment: BrainAmp DC Amplifier (Brain Products GmbH)
Channels: 29 EEG + 2 EOG (horizontal and vertical)
Electrode positions: Standard 10-20 system
Reference: Right mastoid
Sampling rate: 250 Hz
Hardware filter: 0.1 Hz high-pass
Display: 24” TFT, 70 cm viewing distance

Experimental Procedure

Binary communication task: attend left (red cross) for “no”, attend right (green cross) for “yes”
120 statements presented, subjects respond by covert attention shift
Each trial: 10 visual stimuli presentations
Stimulus parameters tested:
- Four symbol sizes: 0.45, 0.90, 1.36, 1.81 degrees visual angle
- Five eccentricities: 4, 5.5, 7, 8.5, 10 degrees visual angle
Inter-stimulus interval: ~175 ms
Online accuracy: 88.5% (+/- 7.8%)

Event Codes

For P300 paradigm compatibility, events are named Target/NonTarget:

NonTarget (1): Left attention (no response)
Target (2): Right attention (yes response)

Data Organization

1 session per subject
120 trials per subject, each with 10 stimulus presentations
Trial duration: 16 seconds (4000 samples at 250 Hz)
Data stored in MAT format with fields:
- bciexp.data: EEG data (channels x samples x trials)
- bciexp.heog, bciexp.veog: EOG data
- bciexp.intention: subject’s intended response (yes/no)
- subject: demographic information

References

[1]

Reichert, C., Tellez-Ceja, I. F., Schwenker, F., Rusnac, A.-L., Curio, G., Aust, L., & Hinrichs, H. (2020). Impact of Stimulus Features on the Performance of a Gaze-Independent Brain-Computer Interface Based on Covert Spatial Attention Shifts. Frontiers in Neuroscience, 14, 591777. https://doi.org/10.3389/fnins.2020.591777

Notes

Added in version 1.3.0.

This dataset uses a covert spatial attention paradigm with N2pc ERP detection, which is different from traditional P300 or motor imagery paradigms. The paradigm is designed for gaze-independent BCI control, making it suitable for users who cannot control eye movements.