moabb.datasets.BNCI2020_002#

class moabb.datasets.BNCI2020_002[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 years

Equipment

  • Amplifier: BrainAmp

  • Electrodes: Ag/AgCl

  • Montage: 10-20

  • Reference: right mastoid

Preprocessing

  • Data state: segmented into trials

  • Bandpass filter: 1-12.5 Hz

  • Steps: highpass filtering

  • Re-reference: Car

Data Access

  • DOI: 10.3389/fnins.2020.591777

Experimental Protocol

  • Paradigm: p300

  • Feedback: visual (yes/no decoded answer)

  • Stimulus: rsvp

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.

See also

BNCI2015_009

AMUSE auditory spatial P300 paradigm

BNCI2015_010

RSVP visual P300 paradigm

Examples

>>> from moabb.datasets import BNCI2020_002
>>> dataset = BNCI2020_002()
>>> dataset.subject_list
[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18]
On this page