NM000108: emg dataset, 20 subjects#

HySER: High-Density Surface Electromyogram Recordings

Access recordings and metadata through EEGDash.

Citation: Xinyu Jiang, Chenyun Dai, Xiangyu Liu, Jiahao Fan (20). HySER: High-Density Surface Electromyogram Recordings. 10.82901/nemar.nm000108

Modality: emg Subjects: 20 Recordings: 1514 License: ODC-By-1.0 Source: nemar

Metadata: Complete (90%)

Quickstart#

Install

pip install eegdash

Access the data

from eegdash.dataset import NM000108

dataset = NM000108(cache_dir="./data")
# Get the raw object of the first recording
raw = dataset.datasets[0].raw
print(raw.info)

Filter by subject

dataset = NM000108(cache_dir="./data", subject="01")

Advanced query

dataset = NM000108(
    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{nm000108,
  title = {HySER: High-Density Surface Electromyogram Recordings},
  author = {Xinyu Jiang and Chenyun Dai and Xiangyu Liu and Jiahao Fan},
  doi = {10.82901/nemar.nm000108},
  url = {https://doi.org/10.82901/nemar.nm000108},
}

About This Dataset#

DOI Paper DOI PhysioNet License

HySER: High-Density Surface Electromyogram Recordings

BIDS-formatted version of the Hyser Surface EMG for Hand Gesture Recognition dataset (Jiang et al., 2021). 20 subjects performed 5 task types across 2 sessions using 256-channel high-density surface EMG (HD-sEMG) with simultaneous 5-finger force recordings.

Subjects

View full README

DOI Paper DOI PhysioNet License

HySER: High-Density Surface Electromyogram Recordings

BIDS-formatted version of the Hyser Surface EMG for Hand Gesture Recognition dataset (Jiang et al., 2021). 20 subjects performed 5 task types across 2 sessions using 256-channel high-density surface EMG (HD-sEMG) with simultaneous 5-finger force recordings.

Subjects

20 right-handed participants (12M, 8F; age 21-34). Two sessions per subject separated by 3-25 days. Demographics in participants.tsv.

Tasks

| Task | BIDS label | Description | Trials |
|------|-----------|-------------|--------|
| Pattern Recognition | `gesture01`-`gesture34` | 34 discrete hand gestures (Table I in paper) | 2 per gesture, each with 3 dynamic + 1 maintenance |
| Maximum Voluntary Contraction | `mvc` | MVC flexion/extension per finger | 2 per finger, 10s each |
| Single Finger (1-DOF) | `singlefinger` | Triangle force trajectory, individual fingers | 3 per finger, 25s each |
| Multi-Finger (N-DOF) | `multifinger` | Simultaneous multi-finger force tracking | 2 per combination, 25s each |
| Random | `random` | Free finger contractions at any force | 5 trials, 25s each |

Equipment

  • EMG system: Quattrocento (OT Bioelettronica, Torino, Italy), 2048 Hz, gain 150, 16-bit ADC

  • Electrodes: Four 8x8 gelled elliptical arrays (5mm x 2.8mm), 10mm inter-electrode distance

  • Placement: Two arrays on extensors (distal/proximal), two on flexors (distal/proximal) of right forearm

  • Reference: Olecranon (elbow); Ground: head of ulna (right leg drive)

  • Hardware filters: HP 10 Hz (2nd order), LP 500 Hz

  • Force sensors: SAS + HSGA (Huatran, Shenzhen, China), 100 Hz, 5 fingers

File Organization

  • *_emg.bdf - 256-channel EMG data (BDF format)

  • *_physio.tsv.gz - 5-finger force data (non-PR tasks only)

  • *_channels.tsv - Channel metadata with electrode mapping (signal_electrode column)

  • *_electrodes.tsv - Electrode positions in local grid coordinates (mm)

  • *_events.tsv - Event markers (gesture trials or segment boundaries)

Non-PR tasks (MVC, singlefinger, multifinger, random) are merged from multiple original recordings into single files per session, with boundary events marking segment junctions.

Coordinate Systems

Four local grid systems (space-ed, space-ep, space-fd, space-fp) in mm, anchored to a parent forearm system (space-forearm) in anatomical percent coordinates. See space-*_coordsystem.json files.

Missing Data

6 gesture recordings absent in source dataset (not conversion failures): sub-01/ses-2/gesture25, sub-03/ses-1/gesture04, sub-03/ses-2/gesture04, sub-05/ses-1/gesture34, sub-11/ses-1/gesture08, sub-19/ses-2/gesture11.

Conversion

Converted using EMG-2-BIDS (EEGLAB + bids-matlab-tools). Data integrity verified: mean Pearson correlation >0.9999 between original WFDB and converted BDF across all 1514 recordings. See dataset_description.json for generator details.

Dataset Information#

Dataset ID

NM000108

Title

HySER: High-Density Surface Electromyogram Recordings

Author (year)

Jiang2021

Canonical

HySER, Hyser

Importable as

NM000108, Jiang2021, HySER, Hyser

Year

20

Authors

Xinyu Jiang, Chenyun Dai, Xiangyu Liu, Jiahao Fan

License

ODC-By-1.0

Citation / DOI

10.82901/nemar.nm000108

Source links

OpenNeuro | NeMAR | Source URL
Copy-paste BibTeX
@dataset{nm000108,
  title = {HySER: High-Density Surface Electromyogram Recordings},
  author = {Xinyu Jiang and Chenyun Dai and Xiangyu Liu and Jiahao Fan},
  doi = {10.82901/nemar.nm000108},
  url = {https://doi.org/10.82901/nemar.nm000108},
}

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: 20

  • Recordings: 1514

  • Tasks: 38

Channels & sampling rate

  • Channels: 256

  • Sampling rate (Hz): Varies

  • Duration (hours): Not calculated

Tags

  • Pathology: Not specified

  • Modality: —

  • Type: —

Files & format

  • Size on disk: 108.2 GB

  • File count: 1514

  • Format: BIDS

License & citation

  • License: ODC-By-1.0

  • DOI: 10.82901/nemar.nm000108

Provenance

API Reference#

Use the NM000108 class to access this dataset programmatically.

class eegdash.dataset.NM000108(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#

Bases: EEGDashDataset

HySER: High-Density Surface Electromyogram Recordings

Study:

nm000108 (NeMAR)

Author (year):

Jiang2021

Canonical:

HySER, Hyser

Also importable as: NM000108, Jiang2021, HySER, Hyser.

Modality: emg. Subjects: 20; recordings: 1514; tasks: 38.

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/nm000108 NeMAR dataset: https://nemar.org/dataexplorer/detail?dataset_id=nm000108 DOI: https://doi.org/10.82901/nemar.nm000108

Examples

>>> from eegdash.dataset import NM000108
>>> dataset = NM000108(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#