NM000189: eeg dataset, 10 subjects#

Name: BNCI 2015-003 P300 dataset
Published: 2011-01-01
License: CC-BY-NC-ND-4.0

BNCI 2015-003 P300 dataset

Access recordings and metadata through EEGDash.

Citation: Martijn Schreuder, Thomas Rost, Michael Tangermann (2011). BNCI 2015-003 P300 dataset.

Modality: eeg Subjects: 10 Recordings: 20 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 NM000189

dataset = NM000189(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 = NM000189(cache_dir="./data", subject="01")

Advanced query

dataset = NM000189(
    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{nm000189,
  title = {BNCI 2015-003 P300 dataset},
  author = {Martijn Schreuder and Thomas Rost and Michael Tangermann},
}

About This Dataset#

BNCI 2015-003 P300 dataset

BNCI 2015-003 P300 dataset.

Dataset Overview

Code: BNCI2015-003
Paradigm: p300
DOI: 10.1016/j.neulet.2009.06.045

View full README

BNCI 2015-003 P300 dataset

BNCI 2015-003 P300 dataset.

Dataset Overview

Code: BNCI2015-003
Paradigm: p300
DOI: 10.1016/j.neulet.2009.06.045
Subjects: 10
Sessions per subject: 1
Events: Target=2, NonTarget=1
Trial interval: [0, 0.8] s
Runs per session: 2
Session IDs: Session 1, Session 2
File format: gdf
Data preprocessed: True
Number of contributing labs: 1

Acquisition

Sampling rate: 256.0 Hz
Number of channels: 8
Channel types: eeg=8
Channel names: Fz, Cz, P3, Pz, P4, PO7, Oz, PO8
Montage: standard_1005
Hardware: BrainAmp
Software: Matlab
Reference: nose
Sensor type: Ag/AgCl electrodes
Line frequency: 50.0 Hz
Online filters: hardware analog band-pass filter between 0.1 and 250 Hz
Impedance threshold: 15.0 kOhm
Cap manufacturer: Brain Products
Electrode type: Ag/AgCl
Electrode material: silver/silver chloride
Auxiliary channels: EOG (2 ch, bipolar)

Participants

Number of subjects: 10
Health status: patients
Clinical population: Healthy
Age: mean=34.1, std=11.4, min=20, max=57
BCI experience: naive
Species: human

Experimental Protocol

Paradigm: p300
Task type: auditory_oddball
Number of classes: 2
Class labels: Target, NonTarget
Tasks: spelling, auditory_attention
Study design: Auditory Multi-class Spatial ERP (AMUSE) paradigm using spatial auditory cues from six speaker locations in azimuth plane. Two-step hex-o-spell like interface for character selection. Subjects mentally count target stimuli from one of six spatial directions.
Study domain: communication
Feedback type: auditory
Stimulus type: spatial_auditory
Stimulus modalities: auditory
Primary modality: auditory
Synchronicity: synchronous
Mode: online
Training/test split: True
Instructions: Focus attention to one target direction and mentally count the number of appearances
Stimulus presentation: soa_ms=175, stimulus_duration_ms=40, stimulus_intensity_db=58, speaker_arrangement=6 speakers at ear height, evenly distributed in circle with 60° distance, radius 65 cm

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: 6
Stimulus onset asynchrony: 175.0 ms

Data Structure

Trials: 48
Trials per class: calibration_per_direction=8
Trials context: calibration_phase

Preprocessing

Data state: filtered
Preprocessing applied: True
Steps: low-pass filter, downsampling, baselining
Highpass filter: 0.1 Hz
Lowpass filter: 40.0 Hz
Bandpass filter: {‘low_cutoff_hz’: 0.1, ‘high_cutoff_hz’: 40.0}
Filter type: analog hardware filter for acquisition; low-pass for online
Artifact methods: variance criterium, peak-to-peak difference criterium
Re-reference: nose
Downsampled to: 100.0 Hz
Epoch window: [-0.15, None]
Notes: For online use signal was low-pass filtered below 40 Hz and downsampled to 100 Hz. Data baselined using 150 ms pre-stimulus data as reference.

Signal Processing

Classifiers: LDA, linear binary classifier
Feature extraction: spatio-temporal features, r2 coefficient, interval averaging
Spatial filters: shrinkage regularization (Ledoit-Wolf)

Cross-Validation

Method: online
Evaluation type: online

Performance (Original Study)

Accuracy: 77.4%
Itr: 2.84 bits/min
Char Per Min Session1: 0.59
Char Per Min Session2 Max: 1.41
Char Per Min Session2 Avg: 0.94
Itr Session2 Avg: 5.26
Itr Session2 Max: 7.55
Success Rate Session1: 76.0

BCI Application

Applications: speller, communication
Environment: laboratory
Online feedback: True

Tags

Pathology: Healthy
Modality: Auditory
Type: ERP, P300

Documentation

Description: Auditory BCI speller using spatial cues (AMUSE paradigm) allowing purely auditory communication interface
DOI: 10.1016/j.neulet.2009.06.045
Associated paper DOI: 10.3389/fnins.2011.00112
License: CC-BY-NC-ND-4.0
Investigators: Martijn Schreuder, Thomas Rost, Michael Tangermann
Senior author: Michael Tangermann
Contact: schreuder@tu-berlin.de
Institution: Berlin Institute of Technology
Department: Machine Learning Laboratory
Address: Machine Learning Laboratory, Berlin Institute of Technology, FR6-9, Franklinstraße 28/29, 10587 Berlin, Germany
Country: Germany
Repository: BNCI Horizon
Publication year: 2011
Funding: European ICT Programme Project FP7-224631; European ICT Programme Project FP7-216886; Deutsche Forschungsgemeinschaft (DFG MU 987/3-2); Bundesministerium fur Bildung und Forschung (BMBF FKZ 01IB001A, 01GQ0850); FP7-ICT PASCAL2 Network of Excellence ICT-216886
Ethics approval: Ethics Committee of the Charité University Hospital
Acknowledgements: Thomas Denck, David List and Larissa Queda for help with experiments. Klaus-Robert Müller and Benjamin Blankertz for fruitful discussions.
Keywords: brain-computer interface, directional hearing, auditory event-related potentials, P300, N200, dynamic subtrials

External Links

Source: http://www.frontiersin.org/neuroprosthetics/10.3389/fnins.2011.00112/abstract

Abstract

This online study introduces an auditory spelling interface that eliminates the necessity for visual representation. In up to two sessions, a group of healthy subjects (N=21) was asked to use a text entry application, utilizing the spatial cues of the AMUSE paradigm (Auditory Multi-class Spatial ERP). The speller relies on the auditory sense both for stimulation and the core feedback. Without prior BCI experience, 76% of the participants were able to write a full sentence during the first session. By exploiting the advantages of a newly introduced dynamic stopping method, a maximum writing speed of 1.41 char/min (7.55 bits/min) could be reached during the second session (average: 0.94 char/min, 5.26 bits/min).

Methodology

Participants surrounded by six speakers at ear height in circle (60° spacing, 65 cm radius). Each direction associated with unique combination of tone (base frequency + harmonics) and band-pass filtered noise. Two-step hex-o-spell interface for character selection. Session 1: calibration (48 trials, 8 per direction, 15 iterations each) followed by online spelling with 15 fixed iterations. Session 2: calibration followed by online spelling with dynamic stopping method (4-15 iterations). Spatio-temporal feature extraction using r2 coefficient and interval selection (2-4 intervals for early and late components, 112-224 features total). Linear binary classifier with shrinkage regularization (Ledoit-Wolf). Decision making based on median classifier scores across iterations.

References

Schreuder, M., Rost, T., & Tangermann, M. (2011). Listen, you are writing! Speeding up online spelling with a dynamic auditory BCI. Frontiers in neuroscience, 5, 112. https://doi.org/10.3389/fnins.2011.00112 Notes .. note:: BNCI2015_003 was previously named BNCI2015003. BNCI2015003 will be removed in version 1.1. .. versionadded:: 0.4.0 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) NeuroTechX/moabb

Dataset Information#

Dataset ID	`NM000189`
Title	BNCI 2015-003 P300 dataset
Author (year)	`Schreuder2015_P300`
Canonical	—
Importable as	`NM000189`, `Schreuder2015_P300`
Year	2011
Authors	Martijn Schreuder, Thomas Rost, Michael Tangermann
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: 10
Recordings: 20
Tasks: 1

Channels & sampling rate

Channels: 8
Sampling rate (Hz): 256.0
Duration (hours): 1.8684006076388888

Electrode Layout#

Electrode layout — EEG · 8 sensors — 8 channels

Dataset Statistics#

Age distribution (n=10, range 34–34 yr)

30

Channel counts: 8 ch (n=40 recordings)

Sampling frequencies: 256.0 Hz (n=40 recordings)

Total recording duration: 1 h 52 min

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

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.

Files: —

Size: —

Subjects: —

Click to load file structure…

API Reference#

Use the NM000189 class to access this dataset programmatically.

class eegdash.dataset.NM000189(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#

Bases: EEGDashDataset

BNCI 2015-003 P300 dataset

Study:: nm000189 (NeMAR)
Author (year):: Schreuder2015_P300
Canonical:: —

Also importable as: NM000189, Schreuder2015_P300.

Modality: eeg; Experiment type: Attention; Subject type: Healthy. Subjects: 10; recordings: 20; 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/nm000189 NeMAR dataset: https://nemar.org/dataexplorer/detail?dataset_id=nm000189

Examples

>>> from eegdash.dataset import NM000189
>>> dataset = NM000189(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)