Discover datasets

An EEG dataset of imagined speech from 15 healthy participants comparing traditional cue-based and gamified paradigm designs for brain-computer interface applications. The dataset comprises 1,800 trials of four Spanish directional commands recorded at 500 Hz from 24 electrodes using the mBrainTrain Smarting system. This derivative dataset enables systematic evaluation of paradigm design effects on imagined speech decoding performance.

Participants

15

Channels

24 (10-10)

Citations

1

HED

v8.4.0

Size

1.52 GB

Version

v1.0.3

Updated

Aug 18, 2026

HEFMI-ICH is a hybrid EEG-fNIRS motor imagery dataset designed for brain-computer interface applications in intracerebral hemorrhage rehabilitation. The dataset comprises 37 participants (17 healthy controls and 20 ICH patients) performing 2-class hand motor imagery tasks (left and right hand grasping) across multiple sessions. With 32-channel EEG recordings at 256 Hz and 3,330 trials, this dataset supports the development and evaluation of BCI systems for clinical stroke rehabilitation.

Participants

37

Channels

32 (10-10)

Citations

5

HED

v8.4.0

Size

6.42 GB

Version

v1.0.3

Updated

Aug 18, 2026

An open-access electroencephalography dataset of motor imagery brain-computer interface performance under six different distraction conditions. Sixteen healthy participants performed left versus right hand motor imagery while experiencing flickering video, number search tasks, news listening, eyes closed state, vibro-tactile stimulation, or no distraction. The dataset comprises 504 trials per subject recorded at 1000 Hz with 63 EEG channels, including one calibration run without feedback and six feedback runs with distinct distraction manipulations.

Participants

16

Channels

63 (10-05)

Citations

13

HED

v8.4.0

Size

72.1 GB

Version

v1.0.7

Updated

Aug 18, 2026

This dataset comprises scalp EEG recordings from 27 stroke patients performing a lower-limb motor imagery task as part of a multi-paradigm, longitudinal rehabilitation training protocol. Data were collected at Tianjin University to support research on motor-imagery brain-computer interfaces for gait and lower-limb rehabilitation after stroke. The dataset is organized in BIDS format with a single EEG task (task-imagery) across repeated sessions.

Participants

27

Channels

64 (10-10)

Citations

23

HED

v8.4.0

Size

56.7 GB

Version

v1.0.5

Updated

Aug 20, 2026

This dataset comprises EEG and EMG recordings from 15 healthy participants performing emergency braking tasks during simulated driving at 200 Hz sampling rate. Participants drove a virtual racing car following a lead vehicle, with occasional abrupt decelerations requiring immediate emergency braking responses. The study demonstrates the feasibility of predicting driver braking intentions from neural and muscular signals before behavioral response, with applications to neuroergonomics and driving assistance systems.

Participants

15

Channels

59 (10-10)

Citations

81

HED

v8.4.0

Size

14.2 GB

Version

v1.0.2

Updated

Aug 18, 2026

This dataset contains 32-channel EEG recordings from 22 healthy adults performing a visual imagery task involving 10 categories of animals, figures, and objects. Data were collected across two sessions per subject using a Neuracle NeuSenW32 system at 1000 Hz sampling rate, and are formatted according to BIDS for use in brain-computer interface research. The dataset supports evaluation of classification approaches such as CSP and EEGNet for decoding imagined visual categories from EEG signals.

Participants

22

Channels

32 (10-10)

Citations

2

HED

v8.4.0

Size

75.4 GB

Version

v1.0.3

Updated

Aug 20, 2026

The BNCI 2015-009 AMUSE dataset comprises EEG recordings from 21 healthy subjects performing an auditory oddball task using spatial hearing cues for brain-computer interface applications. Participants discriminated target and non-target stimuli presented from five spatially distributed speakers with varying inter-stimulus intervals, yielding preprocessed data suitable for P300-based BCI research and offline classification studies.

Participants

21

Channels

60 (10-10)

Citations

160

HED

v8.4.0

Size

16.5 GB

Version

v1.0.2

Updated

Aug 18, 2026

This dataset comprises EEG recordings from 16 healthy participants performing motion visually evoked potential (mVEP) speller tasks using three different paradigm variants: overt gaze-directed, covert attention-based, and foveal motion center stimulation. The study investigates gaze-independent brain-computer interface communication through motion VEP-based P300 visual speller applications with two-level hierarchical selection from 30 symbols (6 symbols per level), employing 63-channel EEG recordings at 100 Hz sampling rate with preprocessed data including artifact rejection and baseline correction.

Participants

16

Channels

63 (10-10)

Citations

44

HED

v8.4.0

Size

3.79 GB

Version

v1.0.2

Updated

Aug 18, 2026

A longitudinal EEG dataset comprising event-related potentials (ERPs) from 15 healthy participants performing a rapid serial visual presentation (RSVP) face authentication task. The dataset includes 4 sessions per subject collected over 200 days, with 57-channel EEG recordings at 1000 Hz sampling rate. Participants discriminated between self-face and AI-generated faces in a p300-based identity authentication paradigm, yielding approximately 160 target and 6240 non-target trials per session. This derivative dataset was generated from raw EEG data using MOABB 1.5.0 and is organized in BIDS format.

Participants

15

Channels

57 (10-10)

Citations

3

HED

v8.4.0

Size

57.0 GB

Version

v1.0.2

Updated

Aug 18, 2026

A P300-based brain-computer interface dataset comprising EEG recordings from 13 healthy participants performing a visual speller task using covert spatial attention and feature attention mechanisms. The dataset includes data from three speller variants (Hex-o-Spell, Cake Speller, and Center Speller) with 63-channel recordings at 250 Hz sampling rate. Preprocessed data includes downsampling, filtering, and baseline correction. The original study reported 97% accuracy on the Center Speller variant.

Participants

13

Channels

63 (10-05)

Citations

81

HED

v8.4.0

Size

9.27 GB

Version

v1.0.2

Updated

Aug 18, 2026