EEGdash›NeMAR›NM000161
Iss. 161 · 20 subjects · 40 recordings · CC-BY-4.0
Dataset Brief · BNCI 2024-001 Handwritten Character Classification dataset

NM000161: eeg dataset, 20 subjects#

BNCI 2024-001 Handwritten Character Classification dataset

Citation: Markus R. Crell, Gernot R. Müller-Putz (2024). BNCI 2024-001 Handwritten Character Classification dataset. 10.82901/nemar.nm000161

20-participant EEG dataset — BNCI 2024-001 Handwritten Character Classification dataset.

Data & curation Markus R. Crell · Gernot R. Müller-Putz
Year 2024 · Distributed via NeMAR
EEG · 60 ch500 HzBIDS 1.9.0Task · imageryHealthyVisualMotor
Layer 01Study
What was asked
Hypothesis, independent & dependent variables, paradigm, cohort, and the editorial caveats around what the recordings can and cannot answer.
Layer 02Signal · BIDS
What was recorded
Sidecars, channels & electrodes, coordinate system, event semantics, and quality stats from the NEMAR pipeline when available.
Layer 03Training · ML
What you can train on
Recommended access modes — MNE Raw, braindecode windows, PyTorch DataLoader — plus the targets the metadata makes addressable.
§ 01Access · Get started

Quickstart#

Install

pip install eegdash

Access the data

from eegdash.dataset import NM000161

dataset = NM000161(cache_dir="./data")
# Get the raw object of the first recording
raw = dataset.datasets[0].raw
print(raw.info)

Filter by subject

dataset = NM000161(cache_dir="./data", subject="01")

Advanced query

dataset = NM000161(
    cache_dir="./data",
    query={"subject": {"$in": ["01", "02"]}},
)

Iterate recordings

for rec in dataset:
    print(rec.subject, rec.raw.info['sfreq'])

If you use this dataset in your research, please cite the original authors.

BibTeX

@dataset{nm000161,
  title = {BNCI 2024-001 Handwritten Character Classification dataset},
  author = {Markus R. Crell and Gernot R. Müller-Putz},
  doi = {10.82901/nemar.nm000161},
  url = {https://doi.org/10.82901/nemar.nm000161},
}
§ 02Study · The README

About This Dataset#

BNCI 2024-001 Handwritten Character Classification dataset.

Schema: HED 8.4.0 | Browse: https://www.hedtags.org/hed-schema-browser

DOI

BNCI 2024-001 Handwritten Character Classification dataset

letter_a
View full README

DOI

BNCI 2024-001 Handwritten Character Classification dataset

letter_a

     ├─ Sensory-event, Experimental-stimulus, Visual-presentation
     └─ Agent-action
        └─ Write
           ├─ Hand
           └─ Label/a

letter_d

     ├─ Sensory-event, Experimental-stimulus, Visual-presentation
     └─ Agent-action
        └─ Write
           ├─ Hand
           └─ Label/d

letter_e

     ├─ Sensory-event, Experimental-stimulus, Visual-presentation
     └─ Agent-action
        └─ Write
           ├─ Hand
           └─ Label/e

letter_f

     ├─ Sensory-event, Experimental-stimulus, Visual-presentation
     └─ Agent-action
        └─ Write
           ├─ Hand
           └─ Label/f

letter_j

     ├─ Sensory-event, Experimental-stimulus, Visual-presentation
     └─ Agent-action
        └─ Write
           ├─ Hand
           └─ Label/j

letter_n

     ├─ Sensory-event, Experimental-stimulus, Visual-presentation
     └─ Agent-action
        └─ Write
           ├─ Hand
           └─ Label/n

letter_o

     ├─ Sensory-event, Experimental-stimulus, Visual-presentation
     └─ Agent-action
        └─ Write
           ├─ Hand
           └─ Label/o

letter_s

     ├─ Sensory-event, Experimental-stimulus, Visual-presentation
     └─ Agent-action
        └─ Write
           ├─ Hand
           └─ Label/s

letter_t

     ├─ Sensory-event, Experimental-stimulus, Visual-presentation
     └─ Agent-action
        └─ Write
           ├─ Hand
           └─ Label/t

letter_v

├─ Sensory-event, Experimental-stimulus, Visual-presentation
└─ Agent-action
   └─ Write
      ├─ Hand
      └─ Label/v

Paradigm-Specific Parameters

  • Detected paradigm: motor_imagery

  • Imagery tasks: handwriting of letters a, d, e, f, j, n, o, s, t, v

  • Cue duration: 2.0 s

  • Imagery duration: 4.0 s

Data Structure

  • Trials: 60

  • Blocks per session: 15

  • Block duration: 340 s

  • Trials context: per_class

Preprocessing

  • Data state: raw

  • Preprocessing applied: True

  • Steps: notch filtering, bandpass filtering, bad channel interpolation, EOG artifact correction (SGEYESUB), ICA for artifact removal, re-referencing to CAR, bad segment rejection, lowpass filtering, downsampling, epoching

  • Highpass filter: 0.3 Hz

  • Lowpass filter: 70.0 Hz

  • Bandpass filter: {‘low_cutoff_hz’: 0.3, ‘high_cutoff_hz’: 70.0}

  • Notch filter: [50] Hz

  • Filter type: Butterworth

  • Filter order: 4

  • Artifact methods: ICA, SGEYESUB

  • Re-reference: car

  • Downsampled to: 128 Hz

  • Epoch window: [-4.5, 4.0]

  • Notes: Two datasets created: dataset 1 (0.3-3 Hz, 10 Hz sampling) and dataset 2 (0.3-40 Hz, 128 Hz sampling). Bad segments rejected if exceeding ±120 μV or kurtosis/probability > 7 SD from mean.

Signal Processing

  • Classifiers: Shrinkage Linear Discriminant Analysis (sLDA), EEGNet CNN

  • Feature extraction: low-frequency EEG, broadband EEG, continuous kinematics decoding

  • Frequency bands: analyzed=[0.3, 70.0] Hz

  • Spatial filters: CAR

Cross-Validation

  • Method: 2-times repeated 5-fold cross-validation

  • Folds: 5

  • Evaluation type: cross_session

Performance (Original Study)

  • Accuracy: 26.2%

  • 10 Letters Direct Lowfreq: 23.1

  • 10 Letters Twostep: 26.2

  • 5 Letters Direct Lowfreq: 39.0

  • 5 Letters Twostep: 46.7

  • Kinematics Correlation Range: 0.10-0.57

  • Chance Level Correlation: 0.04

BCI Application

  • Applications: communication, character_selection

  • Environment: laboratory

  • Online feedback: False

Tags

  • Pathology: Healthy

  • Modality: Motor

  • Type: Motor

Documentation

  • Description: Classification of handwritten letters from EEG through continuous kinematic decoding

  • DOI: 10.1016/j.compbiomed.2024.109132

  • License: CC-BY-4.0

  • Investigators: Markus R. Crell, Gernot R. Müller-Putz

  • Senior author: Gernot R. Müller-Putz

  • Contact: gernot.mueller@tugraz.at

  • Institution: Graz University of Technology

  • Department: Institute of Neural Engineering

  • Address: Graz, Austria

  • Country: Austria

  • Repository: BNCI Horizon 2020

  • Data URL: https://bnci-horizon-2020.eu/database/data-sets

  • Publication year: 2024

  • Ethics approval: Ethics Committee at Graz University of Technology

  • Keywords: Brain-computer interface (BCI), Electroencephalography (EEG), Handwriting, Continuous movement decoding, Non-invasive

Abstract

This study explores the classification of ten letters (a,d,e,f,j,n,o,s,t,v) from non-invasive neural signals of 20 participants. Letters were classified with direct classification from low-frequency and broadband EEG, and a two-step approach comprising continuous decoding of hand kinematics followed by classification. The two-step approach yielded significantly higher performances of 26.2% for ten letters and 46.7% for five letters. Hand kinematics could be reconstructed with correlation of 0.10 to 0.57 (average chance level: 0.04). Results suggest movement speed as the most informative kinematic for decoding short hand movements.

Methodology

Participants wrote 10 letters using right index finger with motion capture tracking (30 Hz, 2D positions). Two-round session with 7 runs (round 1) and 8 runs (round 2), 40 trials per run, 8.5s per trial. Training phase included 4 steps: observation, guided following, unguided following, and execution without feedback. Classification using sliding-window approach with sLDA and EEGNet CNN. Trajectory decoding using EEGNet architecture adapted for regression of position-based (px, py, vx, vy), distance-based (d, ḋ, θ, θ̇), and speed-based (s) kinematics.

References

Crell, M. R., & Muller-Putz, G. R. (2024). Handwritten character classification from EEG through continuous kinematic decoding. Computers in Biology and Medicine, 182, 109132. https://doi.org/10.1016/j.compbiomed.2024.109132 Notes .. versionadded:: 1.3.0 This dataset is notable for exploring non-invasive EEG-based handwritten character classification, which could enable communication for individuals with limited movement capacity. The study demonstrated that handwritten characters can be classified from non-invasive EEG and that decoding movement kinematics prior to classification improves performance.

Appelhoff, S., Sanderson, M., Brooks, T., Vliet, M., Quentin, R., Holdgraf, C., Chaumon, M., Mikulan, E., Tavabi, K., Hochenberger, R., Welke, D., Brunner, C., Rockhill, A., Larson, E., Gramfort, A. and Jas, M. (2019). MNE-BIDS: Organizing electrophysiological data into the BIDS format and facilitating their analysis. Journal of Open Source Software 4: (1896). https://doi.org/10.21105/joss.01896 Pernet, C. R., Appelhoff, S., Gorgolewski, K. J., Flandin, G., Phillips, C., Delorme, A., Oostenveld, R. (2019). EEG-BIDS, an extension to the brain imaging data structure for electroencephalography. Scientific Data, 6, 103. https://doi.org/10.1038/s41597-019-0104-8 Generated by MOABB 1.5.0 (Mother of All BCI Benchmarks) NeuroTechX/moabb

§ 03Cohort · Participants

Cohort#

Dataset Statistics#

Age distribution by gender (n=20, range 28–28 yr, mean 27.0 yr)

25
Other · 20

Channel counts: 60 ch (n=40 recordings)

Sampling frequencies: 500.0 Hz (n=40 recordings)

Total recording duration: 33 h

§ 04Signal · Electrodes & trace

Signal · Electrodes & live trace#

Fig. 01 Signal & montage 60 ch · EEG · 500 Hz · 20 subjects, 40 recordings
Live trace viewer — sub-13 · ses-0 · task-imagery · run-0

Showing one representative recording out of 20 subjects and 40 recordings in this dataset. Browse the full set on OpenNeuro; drop any other _eeg.{set,edf,bdf,vhdr} file onto the viewer (or pass ?eeg=<url>) to inspect it.

Electrode layout — EEG · 60 sensors — 60 channels

NEMAR Processing Statistics#

The plots below are generated by NEMAR’s automated EEG pipeline. The histogram shows pipeline success for data cleaning and ICA decomposition, the percentage of data frames and EEG channels retained after artefact removal, line noise per channel (RMS, dB), and the age/gender distribution of participants.

NEMAR pipeline statistics — NM000161
HED event descriptors word cloud HED event descriptors word cloud — NM000161
§ 05Manifest · BIDS tree

Manifest#

File Explorer#

Browse the BIDS file structure of this dataset. Records are fetched on demand from the EEGDash catalog the first time you open the explorer.

Recordings—
Files—
Subjects—
Modalities—
Click to load file structure…
Full dataset metadata table

Dataset ID

NM000161

Title

BNCI 2024-001 Handwritten Character Classification dataset

Author (year)

Crell2024

Canonical

—

Importable as

NM000161, Crell2024

Year

2024

Authors

Markus R. Crell, Gernot R. Müller-Putz

License

CC-BY-4.0

Citation / DOI

10.82901/nemar.nm000161

Source links

OpenNeuro | NeMAR | Source URL
Copy-paste BibTeX
@dataset{nm000161,
  title = {BNCI 2024-001 Handwritten Character Classification dataset},
  author = {Markus R. Crell and Gernot R. Müller-Putz},
  doi = {10.82901/nemar.nm000161},
  url = {https://doi.org/10.82901/nemar.nm000161},
}
§ 06API · Programmatic access

API Reference#

Signature
eegdash.dataset
class
eegdash.dataset.NM000161(cache_dir, query=None, s3_bucket=None, **kwargs)
Bases: EEGDashDataset
Author (year)Crell2024
Canonical—
Importable asNM000161 · Crell2024
Sourceeegdash/dataset/registry.py · [source ↗]
class eegdash.dataset.NM000161(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#

BNCI 2024-001 Handwritten Character Classification dataset

Study:

nm000161 (NeMAR)

Author (year):

Crell2024

Canonical:

—

Also importable as: NM000161, Crell2024.

Modality: eeg; Experiment type: Motor; Subject type: Healthy. Subjects: 20; recordings: 40; tasks: 1.

Parameters:

  • cache_dir (str | Path) – Directory where data are cached locally.

  • query (dict | None) – Additional MongoDB-style filters to AND with the dataset selection. Must not contain the key dataset.

  • s3_bucket (str | None) – Base S3 bucket used to locate the data.

  • **kwargs (dict) – Additional keyword arguments forwarded to EEGDashDataset.

data_dir#

Local dataset cache directory (cache_dir / dataset_id).

Type:

Path

query#

Merged query with the dataset filter applied.

Type:

dict

records#

Metadata records used to build the dataset, if pre-fetched.

Type:

list[dict] | None

Notes

Each item is a recording; recording-level metadata are available via dataset.description. query supports MongoDB-style filters on fields in ALLOWED_QUERY_FIELDS and is combined with the dataset filter. Dataset-specific caveats are not provided in the summary metadata.

References

OpenNeuro dataset: https://openneuro.org/datasets/nm000161 NeMAR dataset: https://nemar.org/dataexplorer/detail?dataset_id=nm000161 DOI: https://doi.org/10.82901/nemar.nm000161

Examples

>>> from eegdash.dataset import NM000161
>>> dataset = NM000161(cache_dir="./data")
>>> recording = dataset[0]
>>> raw = recording.load()
__init__(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#
save(path: str, overwrite: bool = False, offset: int = 0)[source]#

Save datasets to files by creating one subdirectory for each dataset:

path/
    0/
        0-raw.fif | 0-epo.fif
        description.json
        raw_preproc_kwargs.json (if raws were preprocessed)
        window_kwargs.json (if this is a windowed dataset)
        window_preproc_kwargs.json  (if windows were preprocessed)
        target_name.json (if target_name is not None and dataset is raw)
    1/
        1-raw.fif | 1-epo.fif
        description.json
        raw_preproc_kwargs.json (if raws were preprocessed)
        window_kwargs.json (if this is a windowed dataset)
        window_preproc_kwargs.json  (if windows were preprocessed)
        target_name.json (if target_name is not None and dataset is raw)
Parameters:

  • path (str) –

    Directory in which subdirectories are created to store

    -raw.fif | -epo.fif and .json files to.

  • overwrite (bool) – Whether to delete old subdirectories that will be saved to in this call.

  • offset (int) – If provided, the integer is added to the id of the dataset in the concat. This is useful in the setting of very large datasets, where one dataset has to be processed and saved at a time to account for its original position.

Access modesMNE → braindecode → PyTorch → ML
.rawMNE Raw object — standard tools (filter, epoch, ICA, plot_psd).mne
BaseConcatDatasetEach record is a lazy BaseDataset from braindecode — windowed via create_windows_from_events.braindecode
DataLoaderWraps the windowed dataset into a PyTorch DataLoader; supports parallel workers and on-the-fly augmentations.pytorch
Zarr cacheOptional braindecode Zarr mirror for fast resume; persisted to cache_dir.zarr
Hugging FaceNo per-dataset mirror published yet — browse the EEGDash org listing for sibling datasets. See the datasets loader API.huggingface
Croissant 1.0Machine-readable JSON-LD descriptor — NM000161.croissant.json (MLCommons schema, ingestible by PyTorch / TensorFlow / JAX).mlcommons
Examples using EEGDashcurated · start here
Find datasets with the EEGDash APIQuery the catalogue, filter by task or modality, list candidates.Load one EEG recordingResolve a single record to an MNE Raw with channels and events.EEG recording to PyTorch DataLoaderWrap braindecode windows in a DataLoader for model training.Preprocess EEG and create windowsFilter, resample, epoch — and persist the windowed dataset.Save and reload prepared dataCache a windowed dataset to disk and reattach it without recompute.Download a dataset locallyPrefetch BIDS files to a local cache and validate the layout.

Swap any load_dataset(...) call for nm000161 to reproduce the tutorial on this dataset.

Citation

Markus R. Crell, Gernot R. Müller-Putz (2024). BNCI 2024-001 Handwritten Character Classification dataset. 10.82901/nemar.nm000161

Provenance

¹Contributed to nemar in BIDS format.

²Curated & ingested by the EEGDash catalog; see CITATION.cff for canonical reference.

³Persistent identifier: 10.82901/nemar.nm000161.

BIDS
BIDS 1.9.0
Sidecars
events · events.json · channels · eeg.json
Provenance
CC-BY-4.0 · 10.82901/nemar.nm000161
Machine-readable
schema.org/Dataset · Croissant
Mirrors
OpenNeuro · NEMAR · HF org

See Also#