Discover datasets

A multi-paradigm motor-imagery EEG dataset recorded from healthy participants at Inner Mongolia University. It covers five body-part tasks (hand, foot, thumb, index finger, index-thumb pinch) with left/right laterality, performed under a Classic Arrow paradigm and a Cue-Execution dual-stage paradigm. This BIDS release is a public five-subject sample of a 244-participant dataset, recorded with a 64-channel Neuroscan system at 1000 Hz.

Participants

5

Channels

64 (10-10)

HED

v8.4.0

Size

3.25 GB

Version

v1.0.0

Updated

Oct 8, 2026

A closed-loop EEG dataset of one-dimensional left/right motor imagery cursor control recorded from 10 healthy participants using a 16-channel dry-electrode system. Participants imagined left- or right-hand movement to drive a cursor toward a target under continuous visual feedback, completing 11 runs of 24 trials each while online control parameters were varied. The data support research on sensorimotor rhythm-based brain-computer interfaces and closed-loop decoding.

Participants

10

Channels

16 (other)

HED

v8.4.0

Size

879 MB

Version

v1.0.0

Updated

Oct 8, 2026

EEG recordings of kinesthetic motor imagery of a right-hand ball squeeze at three graded force levels (10%, 40% and 70% of maximal voluntary contraction) from 10 healthy participants. Data were acquired with a 14-channel Emotiv EPOC headset at 128 Hz, with five 5-second imagery cues per 40-second recording. The dataset supports characterization and classification of motor imagery force levels for brain-computer interface research. The associated paper (Journal of Neural Engineering, 2024) is by Martinez-Peon, Garcia-Hernandez, Benavides-Bravo and Parra-Vega.

Participants

10

Channels

14 (10-10)

HED

v8.4.0

Size

54.7 MB

Version

v1.0.0

Updated

Oct 8, 2026

A longitudinal EEG dataset of two-class (left vs. right hand) motor-imagery brain-computer interface training combined with transcutaneous electrical spinal cord stimulation. It includes 27 participants (25 able-bodied and 2 with spinal cord injury), recorded with 32 EEG channels and 3 EOG channels at 512 Hz. The dataset supports study of how spinal stimulation affects sensorimotor activation and BCI skill learning.

Participants

27

Channels

32 (10-10)

HED

v8.4.0

Size

3.53 GB

Version

v1.0.0

Updated

Oct 8, 2026

Pre-execution, two-class hand motor-imagery EEG (precision pinch vs closed fist) recorded with 63 BrainVision channels at 1000 Hz. The dataset includes 10 chronic middle cerebral artery (MCA) stroke patients, recorded at baseline and after rehabilitation, and 8 healthy controls. It supports investigation of mu and beta oscillatory changes (event-related desynchronization) during motor preparation and imagery in stroke rehabilitation.

Participants

18

Channels

59 (10-10)

HED

v8.4.0

Size

5.05 GB

Version

v1.0.0

Updated

Oct 8, 2026

A mobile EEG motor imagery dataset from two pilots, one tetraplegic and one able-bodied, recorded during development of the neuroTUM brain-computer interface for the 2024 Cybathlon BCI race. Data were acquired with a 24-channel wireless Smarting mobi system at 250 Hz using an arrow-cue paradigm with subject-specific sets of three mental tasks (rest, hand and/or foot motor imagery). The dataset is re-released in BIDS format with HED annotations via MOABB.

Participants

2

Channels

24 (10-10)

HED

v8.4.0

Size

408 MB

Version

v1.0.0

Updated

Oct 8, 2026

This dataset contains 63-channel EEG recorded at 512 Hz from 15 stroke patients performing cue-based left-hand versus right-hand motor imagery, with 40 trials per class. For each patient one hand is paretic and the other unaffected, and both were imagined. It is a BIDS re-release of the original figshare deposit, prepared via MOABB with HED event annotations, and is intended for brain-computer interface and neurorehabilitation research.

Participants

15

Channels

63 (other)

HED

v8.4.0

Size

1.74 GB

Version

v1.0.0

Updated

Oct 8, 2026

A longitudinal EEG dataset of 19 healthy right-handed participants trained over four sessions on separate days to control a brain-computer interface. Participants used sustained right-hand grasping motor imagery versus rest to move a cursor on a one-dimensional two-target task with online visual feedback. Each session comprised six feedback runs, giving 14,431 trials in total. The recordings were made with a 74-channel EEG system alongside MEG and are organized in BIDS format.

Participants

19

Channels

74 (10-10)

HED

v8.4.0

Size

3.62 GB

Version

v1.0.0

Updated

Oct 8, 2026

Open-loop EEG dataset of kinesthetic motor imagery of ankle dorsiflexion/plantarflexion alternated with relaxation, recorded from six healthy participants. Each participant completed two sessions, one with a static and one with a moving lower-limb exoskeleton. The recordings contain 28 EEG channels and 4 EOG channels sampled at 250 Hz. This is a BIDS re-release of the Zenodo deposit, produced via MOABB.

Participants

6

Channels

28 (10-10)

HED

v8.4.0

Size

2.88 GB

Version

v1.0.0

Updated

Oct 8, 2026

EEG recordings of motor imagery and motor execution for six activities (raising and lowering each hand, standing and sitting) from 30 healthy participants over two sessions. Data were acquired with a 14-channel Emotiv EPOC X headset at 128 Hz. This BIDS re-release, produced via MOABB, exposes a 3-class imagery task (left hand, right hand, trunk) for brain-computer interface research. The distributed imagery files carry no per-repetition activity label, so only folder-level 3-class labels are available.

Participants

30

Channels

14 (10-10)

HED

v8.4.0

Size

41.2 MB

Version

v1.0.0

Updated

Oct 8, 2026