Discover datasets

EEG recordings from 20 healthy participants who wrote ten letters (a, d, e, f, j, n, o, s, t, v) with the right index finger after a visual cue. The dataset supports classification of handwritten characters, either directly from low-frequency and broadband EEG or with a two-step approach that first decodes continuous hand kinematics and then classifies the letter. It is intended for research on non-invasive brain-computer interfaces for communication.

Participants

20

Channels

60 (10-10)

Citations

13

HED

v8.4.0

Size

27.0 GB

Version

v1.0.2

Updated

Aug 18, 2026

EEG motor imagery dataset recorded from 50 acute stroke patients (1-30 days post-stroke) using a wireless portable semi-dry saline EEG system. Patients imagined left- or right-hand grasping during video-guided trials (40 per subject). It is presented as the first open dataset addressing left- and right-handed motor imagery in acute stroke, supporting BCI algorithm development and rehabilitation applications. Raw and preprocessed data and patient clinical characteristics are provided.

Participants

50

Channels

29 (10-10)

Citations

88

HED

v8.4.0

Size

1.15 GB

Version

v1.0.3

Updated

Aug 18, 2026

EEG recordings from 12 healthy, BCI-naive participants performing kinesthetic motor imagery of upper-limb, elbow-centered tasks. The paradigm has 10 classes: rest, elbow flexion, and eight goal-directed daily-living tasks such as drawer opening, soup eating, weight lifting and combing. Data were recorded with a 17-channel EGI system at 1000 Hz. The dataset supports research on the versatility of EEG models for motor-imagery BCI control and rehabilitation.

Participants

12

Channels

17 (other)

Citations

12

HED

v8.4.0

Size

4.18 GB

Version

v1.0.3

Updated

Aug 18, 2026

M3CV is a multi-subject, multi-session, multi-task EEG database with 64-channel recordings from 95 healthy young adults. Each subject was recorded on two different days across six paradigms covering 13 distributed signal types: resting states, motor execution, transient visual, auditory and somatosensory responses, SSVEP, P300, SSAEP and SSSEP. The data were originally released for an EEG-based biometrics competition on Kaggle. This BIDS version reconstructs pseudo-continuous EEG from the distributed pre-epoched, preprocessed 250 Hz files.

Participants

95

Channels

64 (10-10)

Citations

31

Size

21.6 GB

Version

v1.0.0

Updated

May 11, 2026

This dataset contains 32-channel EEG recorded with a BioSemi ActiveTwo system from 12 healthy participants using a single-flicker, spatially coded SSVEP brain-computer interface for spatial navigation. A single 15 Hz flickering square is shown at the screen center, and four non-flickering targets lie at the cardinal directions. Gaze direction produces distinct spatial topographies of the SSVEP response. Two sessions per subject are included: a structured training calibration session and an online adaptive BCI game session.

Participants

12

Channels

32 (10-10)

Citations

52

HED

v8.4.0

Size

10.9 GB

Version

v1.0.2

Updated

Aug 18, 2026

Stereo-electroencephalography (sEEG) recordings from two epilepsy patients performing bilingual speech tasks: reading Mandarin Chinese, English, and digits aloud. Signals were recorded at 1000 Hz with depth electrodes and are provided as preprocessed derivatives (low-frequency signal, high-gamma activity, and broadband signal), converted to BIDS-iEEG format. The dataset supports research on brain-to-speech decoding and bilingual speech neural activity.

Participants

2

Size

1.87 GB

Version

v1.0.0

Updated

May 4, 2026

This dataset is Study F of the bigP3BCI collection, a P300-based brain-computer interface speller dataset. It contains 16-channel EEG recorded at 256 Hz over three sessions per subject from 10 participants using a 6x6 character grid with multiple stimulus paradigms. Target and non-target events are annotated with HED tags. The data is a BIDS-formatted derivative of the PhysioNet bigP3BCI release, prepared through MOABB.

Participants

10

Channels

16 (10-10)

Citations

2

HED

v8.4.0

Size

1.91 GB

Version

v1.0.2

Updated

Aug 18, 2026

A 40-class SSVEP speller EEG dataset recorded from 40 healthy participants using beta-range flicker frequencies (14.0-21.8 Hz, 0.2 Hz spacing). Stimuli were presented with the joint frequency-phase modulation (JFPM) approach in a 5x8 matrix on a 120 Hz monitor. Each participant completed six blocks of 40 trials with 33-channel BioSemi recordings at 1024 Hz. The dataset is intended for developing low-visual-fatigue SSVEP-based brain-computer interfaces.

Participants

40

Channels

31 (10-10)

Citations

3

HED

v8.4.0

Size

16.5 GB

Version

v1.0.2

Updated

Aug 18, 2026

A benchmark SSVEP-based brain-computer interface dataset containing 64-channel EEG from 34 healthy participants (35 in the original publication) who gazed at 40 flickering characters in a speller. The targets were coded with joint frequency and phase modulation (8-15.8 Hz, 0.2 Hz steps). Each subject completed 6 blocks of 40 cued trials, 5 s of stimulation per trial. Epochs were extracted around stimulus onset and downsampled to 250 Hz.

Participants

34

Channels

64 (10-10)

Citations

98

HED

v8.4.0

Size

6.48 GB

Version

v1.0.3

Updated

Oct 8, 2026

This dataset contains the steady-state visual evoked potential (SSVEP) paradigm of the Mobile BCI dataset from Lee et al. (2021). It provides scalp EEG and EOG recordings from healthy participants who viewed visual flickers at 5.45, 8.57, and 12 Hz while standing, walking, and running on a treadmill. It supports research on brain-computer interface decoding during locomotion.

Participants

23

Channels

73 (other)

Citations

36

HED

v8.4.0

Size

1.31 GB

Version

v1.0.2

Updated

Aug 18, 2026