Discover datasets

A collection of over 1500 one- and two-minute 64-channel EEG recordings from 109 volunteers performing motor execution and motor imagery tasks. Participants completed baseline runs (eyes open and eyes closed) and runs in which they opened and closed, or imagined opening and closing, the left or right fist or both fists or both feet in response to on-screen targets. The data were recorded with the BCI2000 system and are suited to brain-computer interface research.

Participants

109

Channels

64 (10-10)

Citations

26

HED

v8.1.0

Size

11.1 GB

Version

v2.0.0

Updated

Oct 31, 2026

SparrKULee is a speech-evoked auditory response repository from KU Leuven containing EEG recorded from normal-hearing native Dutch/Flemish speakers listening to continuous speech stimuli. This NEMAR re-host contains 80 of the original 85 participants across 11 session types (shortstories01 and varyingStories01-10), plus resting-state recordings. It supports research on neural tracking of speech and auditory EEG modelling.

Participants

80

Channels

64 (10-10)

Citations

49

Size

244 GB

Version

v1.1.1

Updated

Oct 16, 2026

A 128-channel EEG dataset from 10 healthy participants performing two-class (left hand vs. right hand) motor imagery, cued by visual arrows. Each participant completed one session of 150 trials per class. Data were recorded at 500 Hz with a BrainAmp system in a shielded room and re-referenced to common average offline. The dataset was originally collected to evaluate beamforming and spatial filtering approaches for noninvasive brain-computer interfaces. It was converted to BIDS using MOABB (Mother of All BCI Benchmarks) 1.5.0.

Participants

10

Channels

128 (other)

Citations

74

HED

v8.4.0

Size

12.7 GB

Version

v1.0.3

Updated

Oct 16, 2026

A 60-channel EEG dataset from 10 healthy, BCI-naive participants performing simple limb motor imagery (left hand, right hand, feet) and compound limb motor imagery (both hands, left hand with right foot, right hand with left foot), plus a rest condition. The recordings were used to examine EEG oscillatory patterns (ERD/ERS), cognitive processes, and effective connectivity during motor imagery. This is a preprocessed (bandpass filtered, downsampled to 200 Hz) BIDS version curated through MOABB.

Participants

10

Channels

60 (10-10)

Citations

95

HED

v8.4.0

Size

5.64 GB

Version

v1.0.3

Updated

Oct 8, 2026

EEG recordings from 30 healthy participants performing a left-hand grasping motor imagery task versus rest, collected with simultaneous pupillometry. Data were acquired with a 32-channel BioSemi ActiveTwo system at 512 Hz, with 50 auditory-cued trials per participant across two runs. This BIDS derivative was regenerated with a corrected MOABB loader that fixes channel labelling and signal scaling. The dataset supports evaluation of whether task-evoked pupil diameter changes improve EEG-based motor imagery BCI classification.

Participants

30

Channels

32 (10-10)

Citations

38

HED

v8.4.0

Size

3.21 GB

Version

v1.1.0

Updated

Oct 14, 2026

EEG recordings from 10 persons with cervical spinal cord injury performing attempted arm and hand movements: hand open, palmar grasp, lateral grasp, pronation and supination. Sixty-one EEG channels and 3 EOG channels were recorded across 9 runs per participant. Low-frequency time-domain EEG (movement-related cortical potentials) was used to decode the five movement classes offline, and an online proof-of-concept for self-paced movement detection was also run. The dataset supports research on neuroprosthetic and upper-limb control brain-computer interfaces.

Participants

10

Channels

61 (10-05)

Citations

96

HED

v8.4.0

Size

4.51 GB

Version

v1.0.3

Updated

Oct 8, 2026

The eldBETA dataset provides 64-channel EEG recordings from 100 elderly participants (aged 51-81) performing a 9-target SSVEP-based brain-computer interface speller task. Stimuli used joint frequency and phase modulation (JFPM) at 8-12 Hz in 0.5 Hz steps, with 7 blocks of 9 trials per participant. It serves as a benchmark for evaluating SSVEP-BCI decoding algorithms in an aging population.

Participants

100

Channels

64 (10-10)

Citations

28

HED

v8.4.0

Size

34.8 GB

Version

v1.0.4

Updated

Oct 8, 2026

This dataset contains 32-channel EEG recordings from 8 healthy participants performing a combined steady-state visual evoked potential (SSVEP) brain-computer interface task. Spatially overlapping arrow stimuli at a single location, flickering at distinct frequencies derived from an 85 Hz CRT refresh rate, encode four directions (up, down, left, right). Participants covertly attended to one direction per trial, and the data cover three stimulation and attention schemes across two experiments.

Participants

8

Channels

31 (10-10)

Citations

3

HED

v8.4.0

Size

5.30 GB

Version

v1.0.2

Updated

Aug 18, 2026

EEG recordings from nine end-users with central nervous system tissue damage (stroke or spinal cord injury) performing five cue-guided mental tasks: word association, mental subtraction, spatial navigation, right-hand motor imagery, and feet motor imagery. Data were recorded with 30 active EEG electrodes across two sessions per participant, on different days. The dataset supports brain-computer interface research on individually adapted imagery and binary classification of mental tasks.

Participants

9

Channels

30 (10-10)

Citations

83

HED

v8.4.0

Size

2.74 GB

Version

v1.0.2

Updated

Aug 18, 2026

EEG recordings from five healthy participants performing cued motor imagery (right hand versus foot) with 118 electrodes, originally published to demonstrate improved single-trial classification through feature combination and multiclass paradigms. The data were downsampled to 100 Hz and comprise 280 cued trials per subject, split into labeled training and unlabeled test sets. The dataset was part of BCI Competition III and is widely used for benchmarking BCI classification algorithms.

Participants

5

Channels

118 (10-05)

Citations

86

HED

v8.4.0

Size

735 MB

Version

v1.0.3

Updated

Aug 18, 2026