EEGdash›NeMAR›NM000172

Iss. 172 · 14 subjects · 28 recordings · CC-BY-4.0

Dataset Brief · High-gamma dataset described in Schirrmeister et al. 2017

NM000172: eeg dataset, 14 subjects#

High-gamma dataset described in Schirrmeister et al. 2017

Citation: Robin Tibor Schirrmeister, Jost Tobias Springenberg, Lukas Dominique Josef Fiederer, Martin Glasstetter, Katharina Eggensperger, Michael Tangermann, Frank Hutter, Wolfram Burgard, Tonio Ball (2017). High-gamma dataset described in Schirrmeister et al. 2017. 10.82901/nemar.nm000172

14-participant EEG dataset — High-gamma dataset described in Schirrmeister et al. 2017.

Data & curation Robin Tibor Schirrmeister · Jost Tobias Springenberg · Lukas Dominique Josef Fiederer · Martin Glasstetter · Katharina Eggensperger · Michael Tangermann · …
Year 2017 · Distributed via NeMAR
Funding BrainLinks-BrainTools Cluster of Excellence (DFG) EXC1086 · Federal Ministry of Education and Research (BMBF) Motor-BIC 13GW0053D

EEG · 128 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#

Get Started

Install

pip install eegdash

Access the data

from eegdash.dataset import NM000172

dataset = NM000172(cache_dir="./data")
# Get the raw object of the first recording
raw = dataset.datasets[0].raw
print(raw.info)

Query & Filter

Filter by subject

dataset = NM000172(cache_dir="./data", subject="01")

Advanced query

dataset = NM000172(
    cache_dir="./data",
    query={"subject": {"$in": ["01", "02"]}},
)

Iterate recordings

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

Cite This Dataset

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

BibTeX

@dataset{nm000172,
  title = {High-gamma dataset described in Schirrmeister et al. 2017},
  author = {Robin Tibor Schirrmeister and Jost Tobias Springenberg and Lukas Dominique Josef Fiederer and Martin Glasstetter and Katharina Eggensperger and Michael Tangermann and Frank Hutter and Wolfram Burgard and Tonio Ball},
  doi = {10.82901/nemar.nm000172},
  url = {https://doi.org/10.82901/nemar.nm000172},
}

§ 02Study · The README

About This Dataset#

High-gamma dataset described in Schirrmeister et al. 2017.

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

DOI

High-gamma dataset described in Schirrmeister et al. 2017

right_hand

View full README

DOI

High-gamma dataset described in Schirrmeister et al. 2017

right_hand

     ├─ Sensory-event, Experimental-stimulus, Visual-presentation
     └─ Agent-action
        └─ Imagine
           ├─ Move
           └─ Right, Hand

left_hand

     ├─ Sensory-event, Experimental-stimulus, Visual-presentation
     └─ Agent-action
        └─ Imagine
           ├─ Move
           └─ Left, Hand

rest

     ├─ Sensory-event
     ├─ Experimental-stimulus
     ├─ Visual-presentation
     └─ Rest

feet

├─ Sensory-event, Experimental-stimulus, Visual-presentation
└─ Agent-action
   └─ Imagine, Move, Foot

Paradigm-Specific Parameters

Detected paradigm: motor_imagery
Imagery tasks: left_hand_finger_tapping, right_hand_finger_tapping, feet_toe_clenching, rest

Data Structure

Trials: {‘total_per_subject’: 963, ‘training_set’: 880, ‘test_set’: 160}
Trials per class: per_class_per_subject=260
Blocks per session: 13
Trials context: 13 runs per subject, 80 trials per run (4 seconds each), 3-4 seconds inter-trial interval, pseudo-randomized presentation with all 4 classes shown every 4 trials

Signal Processing

Classifiers: Deep ConvNet, Shallow ConvNet, ResNet, FBCSP with LDA
Feature extraction: FBCSP, CSP, Bandpower, Spectral power modulations
Frequency bands: alpha=[7.0, 13.0] Hz; beta=[13.0, 30.0] Hz; gamma=[30.0, 100.0] Hz
Spatial filters: CSP

Cross-Validation

Method: holdout
Evaluation type: within_subject

Performance (Original Study)

Fbcsp Accuracy: 91.2
Deep Convnet Accuracy: 89.3
Shallow Convnet Accuracy: 92.5

BCI Application

Applications: motor_control
Environment: laboratory
Online feedback: False

Tags

Pathology: Healthy
Modality: Motor
Type: Motor Imagery, Motor Execution

Documentation

DOI: 10.1002/hbm.23730
License: CC-BY-4.0
Investigators: Robin Tibor Schirrmeister, Jost Tobias Springenberg, Lukas Dominique Josef Fiederer, Martin Glasstetter, Katharina Eggensperger, Michael Tangermann, Frank Hutter, Wolfram Burgard, Tonio Ball
Senior author: Tonio Ball
Contact: robin.schirrmeister@uniklinik-freiburg.de
Institution: University of Freiburg
Department: Translational Neurotechnology Lab, Epilepsy Center, Medical Center
Address: Engelberger Str. 21, Freiburg 79106, Germany
Country: DE
Repository: GitHub
Data URL: https://web.gin.g-node.org/robintibor/high-gamma-dataset/
Publication year: 2017
Funding: BrainLinks-BrainTools Cluster of Excellence (DFG) EXC1086; Federal Ministry of Education and Research (BMBF) Motor-BIC 13GW0053D
Ethics approval: Approved by the ethical committee of the University of Freiburg
Acknowledgements: Funded by BrainLinks-BrainTools Cluster of Excellence (DFG, EXC1086) and the Federal Ministry of Education and Research (BMBF, Motor-BIC 13GW0053D).
How to acknowledge: Please cite: Schirrmeister et al. (2017). Deep learning with convolutional neural networks for EEG decoding and visualization. Human Brain Mapping, 38(11), 5391-5420. https://doi.org/10.1002/hbm.23730
Keywords: electroencephalography, EEG analysis, machine learning, end-to-end learning, brain-machine interface, brain-computer interface, model interpretability, brain mapping

Abstract

Deep learning with convolutional neural networks (deep ConvNets) has revolutionized computer vision through end-to-end learning. This study investigates deep ConvNets for end-to-end EEG decoding of imagined or executed movements from raw EEG. Results show that recent advances including batch normalization and exponential linear units, together with a cropped training strategy, boosted decoding performance to match or exceed FBCSP (82.1% FBCSP vs 84.0% deep ConvNets). Novel visualization methods demonstrated that ConvNets learned to use spectral power modulations in alpha, beta, and high gamma frequencies with meaningful spatial distributions.

Methodology

End-to-end deep learning approach comparing shallow ConvNets, deep ConvNets, and ResNets against FBCSP baseline. Evaluated design choices including batch normalization, exponential linear units, dropout, and cropped training strategies. Novel visualization techniques developed to understand learned features and verify that ConvNets use spectral power modulations in task-relevant frequency bands.

References

Schirrmeister, Robin Tibor, et al. “Deep learning with convolutional neural networks for EEG decoding and visualization.” Human brain mapping 38.11 (2017): 5391-5420.

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=14, range 27–27 yr, mean 27.0 yr)

25

Other · 14

Channel counts: 128 ch (n=56 recordings)

Sampling frequencies: 500.0 Hz (n=56 recordings)

Total recording duration: 57 h

§ 04Signal · Electrodes & trace

Signal · Electrodes & live trace#

Fig. 01 Signal & montage 128 ch · EEG · 500 Hz · 14 subjects, 28 recordings

Live trace viewer — sub-13 · ses-0 · task-imagery · run-1

Showing one representative recording out of 14 subjects and 28 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 · 128 sensors — 128 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.

HED event descriptors word cloud

§ 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…

§ 06API · Programmatic access

API Reference#

Signature

eegdash.dataset

class

eegdash.dataset.NM000172(cache_dir, query=None, s3_bucket=None, **kwargs)

Bases: EEGDashDataset

Author (year)Schirrmeister2017

Canonical—

Importable asNM000172 · Schirrmeister2017

Sourceeegdash/dataset/registry.py · [source ↗]

class eegdash.dataset.NM000172(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#

High-gamma dataset described in Schirrmeister et al. 2017

Study:: nm000172 (NeMAR)
Author (year):: Schirrmeister2017
Canonical:: —

Also importable as: NM000172, Schirrmeister2017.

Modality: eeg; Experiment type: Motor; Subject type: Healthy. Subjects: 14; recordings: 28; 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/nm000172 NeMAR dataset: https://nemar.org/dataexplorer/detail?dataset_id=nm000172 DOI: https://doi.org/10.82901/nemar.nm000172

Examples

>>> from eegdash.dataset import NM000172
>>> dataset = NM000172(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 — NM000172.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 nm000172 to reproduce the tutorial on this dataset.

Citation

Robin Tibor Schirrmeister, Jost Tobias Springenberg, Lukas Dominique Josef Fiederer, Martin Glasstetter, Katharina Eggensperger, … (2017). High-gamma dataset described in Schirrmeister et al. 2017. 10.82901/nemar.nm000172

Provenance

¹Contributed to nemar in BIDS format.

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

³Persistent identifier: 10.82901/nemar.nm000172.

Related & sibling datasets

NM000145EEG · 10 subj NM000162EEG · 20 subj NM000161EEG · 20 subj NM000140EEG · 12 subj NM000171EEG · 14 subj

+ 1 more — see See Also below →

BIDS

BIDS 1.9.0

Sidecars

events · events.json · channels · eeg.json

Provenance

CC-BY-4.0 · 10.82901/nemar.nm000172

Machine-readable

schema.org/Dataset · Croissant

Mirrors

OpenNeuro · NEMAR · HF org

Dataset ID	`NM000172`
Title	High-gamma dataset described in Schirrmeister et al. 2017
Author (year)	`Schirrmeister2017`
Canonical	—
Importable as	`NM000172`, `Schirrmeister2017`
Year	2017
Authors	Robin Tibor Schirrmeister, Jost Tobias Springenberg, Lukas Dominique Josef Fiederer, Martin Glasstetter, Katharina Eggensperger, Michael Tangermann, Frank Hutter, Wolfram Burgard, Tonio Ball
License	CC-BY-4.0
Citation / DOI	10.82901/nemar.nm000172
Source links	OpenNeuro \| NeMAR \| Source URL

NM000172: eeg dataset, 14 subjects#

Quickstart#

About This Dataset#

Cohort#

Dataset Statistics#

Signal · Electrodes & live trace#

NEMAR Processing Statistics#

Manifest#

File Explorer#

API Reference#

Citation

Provenance

Related & sibling datasets

See Also#