NM000203: eeg dataset, 13 subjects#

P300 dataset from initial spot study

Access recordings and metadata through EEGDash.

Citation: Jan Sosulski, David Hübner, Aaron Klein, Michael Tangermann (2019). P300 dataset from initial spot study.

Modality: eeg Subjects: 13 Recordings: 800 License: CC-BY-SA-4.0 Source: openneuro

Metadata: Complete (90%)

Quickstart#

Install

pip install eegdash

Access the data

from eegdash.dataset import NM000203

dataset = NM000203(cache_dir="./data")
# Get the raw object of the first recording
raw = dataset.datasets[0].raw
print(raw.info)

Filter by subject

dataset = NM000203(cache_dir="./data", subject="01")

Advanced query

dataset = NM000203(
    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{nm000203,
  title = {P300 dataset from initial spot study},
  author = {Jan Sosulski and David Hübner and Aaron Klein and Michael Tangermann},
}

About This Dataset#

P300 dataset from initial spot study

P300 dataset from initial spot study.

Dataset Overview

  • Code: Sosulski2019

  • Paradigm: p300

  • DOI: 10.6094/UNIFR/154576

View full README

P300 dataset from initial spot study

P300 dataset from initial spot study.

Dataset Overview

  • Code: Sosulski2019

  • Paradigm: p300

  • DOI: 10.6094/UNIFR/154576

  • Subjects: 13

  • Sessions per subject: 80

  • Events: Target=21, NonTarget=1

  • Trial interval: [-0.2, 1] s

  • File format: brainvision

Acquisition

  • Sampling rate: 1000.0 Hz

  • Number of channels: 31

  • Channel types: eeg=31, eog=1, misc=5

  • Channel names: C3, C4, CP1, CP2, CP5, CP6, Cz, EOGvu, F10, F3, F4, F7, F8, F9, FC1, FC2, FC5, FC6, Fp1, Fp2, Fz, O1, O2, P10, P3, P4, P7, P8, P9, Pz, T7, T8, x_EMGl, x_GSR, x_Optic, x_Pulse, x_Respi

  • Montage: standard_1020

  • Hardware: BrainProducts BrainAmp DC

  • Reference: nose

  • Sensor type: passive Ag/AgCl

  • Line frequency: 50.0 Hz

  • Auxiliary channels: EOG (1 ch, vertical)

Participants

  • Number of subjects: 13

  • Health status: healthy

  • Age: mean=22.7, std=1.64, min=20, max=26

  • Gender distribution: male=5, female=8

  • Species: human

Experimental Protocol

  • Paradigm: p300

  • Number of classes: 2

  • Class labels: Target, NonTarget

  • Study design: Subjects focused attention on target tones (1000 Hz) and ignored non-target tones (500 Hz) presented via speaker at 65 cm distance. One trial consisted of 15 target and 75 non-target stimuli in pseudo-random order with at least two non-target tones between target tones. The experiment was split into optimization and validation parts.

  • Stimulus type: oddball

  • Stimulus modalities: auditory

  • Primary modality: auditory

  • Synchronicity: synchronous

  • Mode: online

  • Instructions: Focus on the target tones (1000 Hz) and ignore the non-target tones (500 Hz). Refrain from blinking and movement as much as possible.

  • Stimulus presentation: target_tone_hz=1000, non_target_tone_hz=500, tone_duration_ms=40, distance_cm=65

HED Event Annotations

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

Target

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

NonTarget

├─ Sensory-event
├─ Experimental-stimulus
├─ Visual-presentation
└─ Non-target

Paradigm-Specific Parameters

  • Detected paradigm: p300

  • Number of targets: 1

Data Structure

  • Trials: Variable: optimization part used time-limited trials (20 minutes per strategy), validation part used 20 trials per SOA

  • Trials per class: target=13 per trial (after preprocessing, originally 15), non_target=65 per trial (after preprocessing, originally 75)

  • Trials context: Each trial consisted of 90 stimuli (15 target, 75 non-target). After preprocessing (removing first and last 6 epochs), 78 data points available per trial: 13 target and 65 non-target epochs.

Signal Processing

  • Classifiers: rLDA, Shrinkage LDA

  • Feature extraction: Mean amplitude in time intervals

  • Frequency bands: analyzed=[1.5, 40.0] Hz

Cross-Validation

  • Method: 13-fold

  • Folds: 13

  • Evaluation type: within_session

Performance (Original Study)

  • Auc: 0.701

  • Mean Auc Ucb: 0.701

  • Mean Auc Rand: 0.704

  • Mean Auc P300 Ucb: 0.67

  • Mean Auc P300 Rand: 0.681

  • Mean Auc Fixed60: 0.517

BCI Application

  • Applications: communication

  • Online feedback: False

Tags

  • Pathology: Healthy

  • Modality: Auditory

  • Type: Research

Documentation

  • Description: Auditory oddball ERP dataset from 13 healthy subjects. Two sinusoidal tones (target 1000 Hz, non-target 500 Hz) presented at various stimulus onset asynchronies (SOAs, 60-600 ms). 31-channel EEG recorded at 1000 Hz with BrainProducts BrainAmp DC. Raw BrainVision format data.

  • DOI: 10.48550/arXiv.2109.06011

  • License: CC-BY-SA-4.0

  • Investigators: Jan Sosulski, David Hübner, Aaron Klein, Michael Tangermann

  • Senior author: Michael Tangermann

  • Contact: jan.sosulski@blbt.uni-freiburg.de; davhuebn@gmail.com; kleinaa@cs.uni-freiburg.de; michael.tangermann@donders.ru.nl

  • Institution: University of Freiburg

  • Country: DE

  • Repository: FreiDok

  • Data URL: https://freidok.uni-freiburg.de/data/154576

  • Publication year: 2021

  • Funding: Cluster of Excellence BrainLinks-BrainTools funded by the German Research Foundation (DFG) [grant number EXC 1086]; DFG project SuitAble [grant number TA 1258/1-1]; state of Baden-Württemberg, Germany, through bwHPC and the German Research Foundation (DFG) [grant number INST 39/963-1 FUGG]

  • Ethics approval: Approved by the ethics committee of the university medical center of Freiburg

  • Acknowledgements: Experiments were performed according to the Declaration of Helsinki.

  • Keywords: Bayesian optimization, individual experimental parameters, brain-computer interfaces, learning from small data, auditory event-related potentials, closed-loop parameter optimization

Abstract

The decoding of brain signals recorded via, e.g., an electroencephalogram, using machine learning is key to brain-computer interfaces (BCIs). Stimulation parameters or other experimental settings of the BCI protocol typically are chosen according to the literature. The decoding performance directly depends on the choice of parameters, as they influence the elicited brain signals and optimal parameters are subject-dependent. Thus a fast and automated selection procedure for experimental parameters could greatly improve the usability of BCIs. We evaluate a standalone random search and a combined Bayesian optimization with random search into a closed-loop auditory event-related potential protocol. We aimed at finding the individually best stimulation speed—also known as stimulus onset asynchrony (SOA)—that maximizes the classification performance of a regularized linear discriminant analysis.

Methodology

The experiment was divided into two parts: (1) Optimization part: four strategies (AUC-ucb, AUC-rand, P300-ucb, P300-rand) each allocated 20 minutes to find optimal SOA. Strategies alternated to minimize non-stationarity effects. (2) Validation part: evaluated SOAs from each optimization strategy plus fixed 60ms SOA using 20 trials each (in blocks of 5 trials). Features were mean amplitudes in 5 time intervals ([100, 170], [171, 230], [231, 300], [301, 410], [411, 500] ms) across 31 channels (155 dimensions total). Classification used rLDA with automatic shrinkage regularization and 13-fold cross-validation on single trials.

References

Sosulski, J., Tangermann, M.: Electroencephalogram signals recorded from 13 healthy subjects during an auditory oddball paradigm under different stimulus onset asynchrony conditions. Dataset. DOI: 10.6094/UNIFR/154576 Sosulski, J., Tangermann, M.: Spatial filters for auditory evoked potentials transfer between different experimental conditions. Graz BCI Conference. 2019. Sosulski, J., Hübner, D., Klein, A., Tangermann, M.: Online Optimization of Stimulation Speed in an Auditory Brain-Computer Interface under Time Constraints. arXiv preprint. 2021. Notes .. versionadded:: 0.4.5 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

Dataset Information#

Dataset ID

NM000203

Title

P300 dataset from initial spot study

Author (year)

—

Canonical

—

Importable as

NM000203

Year

2019

Authors

Jan Sosulski, David Hübner, Aaron Klein, Michael Tangermann

License

CC-BY-SA-4.0

Citation / DOI

Unknown

Source links

OpenNeuro | NeMAR | Source URL

Found an issue with this dataset?

If you encounter any problems with this dataset (missing files, incorrect metadata, loading errors, etc.), please let us know!

Report an Issue on GitHub

Technical Details#

Subjects & recordings

  • Subjects: 13

  • Recordings: 800

  • Tasks: 1

Channels & sampling rate

  • Channels: 31

  • Sampling rate (Hz): 1000.0

  • Duration (hours): 8.911166666666666

Tags

  • Pathology: Not specified

  • Modality: —

  • Type: —

Files & format

  • Size on disk: 2.8 GB

  • File count: 800

  • Format: BIDS

License & citation

  • License: CC-BY-SA-4.0

  • DOI: —

Provenance

API Reference#

Use the NM000203 class to access this dataset programmatically.

class eegdash.dataset.NM000203(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#

Bases: EEGDashDataset

P300 dataset from initial spot study

Study:

nm000203 (OpenNeuro)

Author (year):

nan

Canonical:

—

Also importable as: NM000203, nan.

Modality: eeg. Subjects: 13; recordings: 800; 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/nm000203 NeMAR dataset: https://nemar.org/dataexplorer/detail?dataset_id=nm000203

Examples

>>> from eegdash.dataset import NM000203
>>> dataset = NM000203(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, overwrite=False)[source]#

Save the dataset to disk.

Parameters:

  • path (str or Path) – Destination file path.

  • overwrite (bool, default False) – If True, overwrite existing file.

Return type:

None

See Also#