NM000194: eeg dataset, 12 subjects

BNCI 2015-010 RSVP P300 dataset

Access recordings and metadata through EEGDash.

Citation: Laura Acqualagna, Benjamin Blankertz (2013). BNCI 2015-010 RSVP P300 dataset.

Modality: eeg Subjects: 12 Recordings: 24 License: CC-BY-NC-ND-4.0 Source: nemar

Metadata: Complete (90%)

Quickstart

Get Started

Install

pip install eegdash

Access the data

from eegdash.dataset import NM000194

dataset = NM000194(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 = NM000194(cache_dir="./data", subject="01")

Advanced query

dataset = NM000194(
    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{nm000194,
  title = {BNCI 2015-010 RSVP P300 dataset},
  author = {Laura Acqualagna and Benjamin Blankertz},
}

About This Dataset

BNCI 2015-010 RSVP P300 dataset

BNCI 2015-010 RSVP P300 dataset.

Dataset Overview

• Code: BNCI2015-010

• Paradigm: p300

• DOI: 10.1016/j.clinph.2012.12.050

View full README

BNCI 2015-010 RSVP P300 dataset

BNCI 2015-010 RSVP P300 dataset.

Dataset Overview

• Code: BNCI2015-010

• Paradigm: p300

• DOI: 10.1016/j.clinph.2012.12.050

• Subjects: 12

• Sessions per subject: 1

• Events: Target=1, NonTarget=2

• Trial interval: [0, 0.8] s

• Runs per session: 2

• Session IDs: calibration, copy-spelling, free-spelling

• File format: EEG

• Data preprocessed: True

Acquisition

• Sampling rate: 200.0 Hz

• Number of channels: 63

• Channel types: eeg=63

• Channel names: Fp1, Fp2, AF3, AF4, Fz, F1, F2, F3, F4, F5, F6, F7, F8, F9, F10, FCz, FC1, FC2, FC3, FC4, FC5, FC6, FT7, FT8, Cz, C1, C2, C3, C4, C5, C6, T7, T8, CPz, CP1, CP2, CP3, CP4, CP5, CP6, TP7, TP8, Pz, P1, P2, P3, P4, P5, P6, P7, P8, P9, P10, POz, PO3, PO4, PO7, PO8, PO9, PO10, Oz, O1, O2

• Montage: 10-20

• Hardware: BrainAmp amplifiers

• Software: Python with Pyff framework

• Reference: left mastoid

• Sensor type: active electrode

• Line frequency: 50.0 Hz

• Online filters: lowpass Chebyshev filter up to 40 Hz

• Impedance threshold: 10.0 kOhm

• Cap manufacturer: Brain Products

• Cap model: actiCap

• Electrode type: active electrode

Participants

• Number of subjects: 12

• Health status: patients

• Clinical population: Healthy

• Age: mean=29.17, std=8.4, min=24, max=55

• Gender distribution: male=6, female=6

• Handedness: all right-handed

• BCI experience: mixed

• Species: human

Experimental Protocol

• Paradigm: p300

• Task type: spelling

• Number of classes: 2

• Class labels: Target, NonTarget

• Trial duration: 46.5 s

• Study design: RSVP (Rapid Serial Visual Presentation) BCI speller where 30 symbols are presented one-by-one in random order at the center of the screen. Three conditions tested: NoColor 116ms SOA, Color 116ms SOA, and Color 83ms SOA. Colors used to facilitate discrimination.

• Feedback type: visual

• Stimulus type: RSVP letters

• Stimulus modalities: visual

• Primary modality: visual

• Synchronicity: synchronous

• Mode: online

• Training/test split: True

• Instructions: Participants fixate center of screen, concentrate on target letter, silently count its occurrences. Avoid blinking during visual presentation.

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: 30

• Number of repetitions: 10

• Stimulus onset asynchrony: 116.0 ms

Data Structure

• Trials: 10 sequences of 30 symbols

• Blocks per session: 3

• Trials context: per sequence

Preprocessing

• Data state: filtered

• Preprocessing applied: True

• Steps: lowpass filter, downsampling, baseline correction, artifact rejection

• Lowpass filter: 40.0 Hz

• Filter type: Chebyshev

• Filter order: passband up to 40 Hz, stopband starting at 49 Hz

• Artifact methods: min-max criterion for eye movement rejection (75 µV on F9, Fz, F10, AF3, AF4), broadband power rejection (5-40 Hz)

• Re-reference: linked mastoids (offline)

• Downsampled to: 200.0 Hz

• Epoch window: [-0.1, 1.2]

• Notes: Baseline correction on pre-stimulus interval (116ms for 116ms SOA, 83/2ms for 83ms SOA). Non-target epochs excluded if 3 preceding or following symbols were targets.

Signal Processing

• Classifiers: LDA with shrinkage

• Feature extraction: spatio-temporal features, averaged voltages within time windows

• Frequency bands: alpha=[7, 13] Hz

• Spatial filters: 55 channels used for classification (all except Fp1,2, AF3,4, F9,10, FT7,8)

Cross-Validation

• Method: calibration/test split

• Evaluation type: within_session

Performance (Original Study)

• Accuracy: 94.8%

• Mean Spelling Rate Symb Per Min: 1.43

• Trial Duration 116Ms Soa S: 46.5

• Trial Duration 83Ms Soa S: 36.6

BCI Application

• Applications: speller, communication

• Environment: laboratory

• Online feedback: True

Tags

• Pathology: Healthy

• Modality: Visual

• Type: ERP

Documentation

• DOI: 10.1016/j.clinph.2012.12.050

• License: CC-BY-NC-ND-4.0

• Investigators: Laura Acqualagna, Benjamin Blankertz

• Senior author: Benjamin Blankertz

• Contact: laura.acqualagna@tu-berlin.de; benjamin.blankertz@tu-berlin.de

• Institution: Berlin Institute of Technology

• Department: Machine Learning Laboratory; Neurotechnology Group

• Country: Germany

• Repository: BNCI Horizon

• Publication year: 2013

• Funding: BMBF Grant; Grant Nos s; Grant No. MU MU; DFG Grant

• Ethics approval: Study performed in accordance with the declaration of Helsinki

• Keywords: Brain Computer Interfaces, RSVP, ERPs, Speller, P300, N2, gaze-independent

Abstract

A Brain Computer Interface (BCI) speller using rapid serial visual presentation (RSVP) paradigm for gaze-independent mental typewriting. Twelve healthy participants successfully operated the RSVP speller with mean online spelling rate of 1.43 symb/min and mean symbol selection accuracy of 94.8%. The RSVP speller does not require gaze shifts and can be operated by non-spatial visual attention, making it suitable for patients with impaired oculo-motor control.

Methodology

Three experimental conditions tested (NoColor 116ms, Color 116ms, Color 83ms SOA). Each condition included calibration, copy-spelling, and free-spelling phases. Vocabulary of 30 symbols presented one-by-one at screen center in pseudo-random order. EEG recorded at 1000 Hz with 63 channels, downsampled to 200 Hz for ERP analysis. Classification using LDA with shrinkage on spatio-temporal features from 5 individually selected time windows. Symbol selection based on averaged classifier output across 10 sequences.

References

Acqualagna, L., & Blankertz, B. (2013). Gaze-independent BCI-spelling using rapid serial visual presentation (RSVP). Clinical Neurophysiology, 124(5), 901-908. https://doi.org/10.1016/j.clinph.2012.12.050 Notes .. versionadded:: 1.2.0 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) https://github.com/NeuroTechX/moabb

Dataset Information

Dataset ID	NM000194
Title	BNCI 2015-010 RSVP P300 dataset
Author (year)	Acqualagna2015
Canonical	BNCI2015
Importable as	NM000194, Acqualagna2015, BNCI2015
Year	2013
Authors	Laura Acqualagna, Benjamin Blankertz
License	CC-BY-NC-ND-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: 12

• Recordings: 24

• Tasks: 1

Channels & sampling rate

• Channels: 63 (22), 61 (2)

• Sampling rate (Hz): 200.0

• Duration (hours): 16.163227777777777

Tags

• Pathology: Healthy

• Modality: Visual

• Type: Attention

Files & format

• Size on disk: 2.1 GB

• File count: 24

• Format: BIDS

License & citation

• License: CC-BY-NC-ND-4.0

• DOI: —

Provenance

• Source: nemar

• OpenNeuro: nm000194

• NeMAR: nm000194

API Reference

Use the NM000194 class to access this dataset programmatically.

class eegdash.dataset.NM000194(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)

Bases: EEGDashDataset

BNCI 2015-010 RSVP P300 dataset

Study:

nm000194 (NeMAR)

Author (year):

Acqualagna2015

Canonical:

—

Also importable as: NM000194, Acqualagna2015.

Modality: eeg; Experiment type: Attention; Subject type: Healthy. Subjects: 12; recordings: 24; 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/nm000194 NeMAR dataset: https://nemar.org/dataexplorer/detail?dataset_id=nm000194

Examples

>>> from eegdash.dataset import NM000194
>>> dataset = NM000194(cache_dir="./data")
>>> recording = dataset[0]
>>> raw = recording.load()

__init__(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)

save(path, overwrite=False)

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

• eegdash.dataset.EEGDashDataset

• eegdash.dataset

• OpenNeuro dataset page

• NeMAR dataset page

• eegdash.dataset.DS001785

• eegdash.dataset.DS001787

• eegdash.dataset.DS001810

• eegdash.dataset.DS001849

• eegdash.dataset.DS001971