EEGdash›NeMAR›NM000413
Iss. 413 · 29 subjects · 190 recordings · CC-BY-4.0
Dataset Brief · DRE-LZSH

NM000413: ieeg dataset, 29 subjects#

DRE-LZSH: SEEG benchmark dataset for drug-resistant epilepsy (ictal and interictal segments, 29 patients)

Access recordings and metadata through EEGDash.

Citation: Wang Yalin (20). DRE-LZSH: SEEG benchmark dataset for drug-resistant epilepsy (ictal and interictal segments, 29 patients). 10.82901/nemar.nm000413

Modality: ieeg Subjects: 29 Recordings: 190 License: CC-BY-4.0 Source: nemar

Metadata: Complete (100%)

29-participant iEEG dataset — DRE-LZSH: SEEG benchmark dataset for drug-resistant epilepsy (ictal and interictal segments, 29 patients).

Data & curation Wang Yalin
Senior author Wang Yalin
Year 20 · Distributed via NeMAR
iEEG · 107 (18), 113 (14), 91 (14), 117 (13), 111 (10), 123 (10), 163 (10), 93 (10), 126 (9), 79 (8), 108 (8), 83 (7), 95 (6), 118 (6), 128 (6), 139 (6), 114 (6), 85 (6), 124 (6), 73 (4), 149 (4), 121 (4), 86 (3), 104 (2) ch2000 HzBIDS 1.10.02 tasks
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#

Install

pip install eegdash

Access the data

from eegdash.dataset import NM000413

dataset = NM000413(cache_dir="./data")
# Get the raw object of the first recording
raw = dataset.datasets[0].raw
print(raw.info)

Filter by subject

dataset = NM000413(cache_dir="./data", subject="01")

Advanced query

dataset = NM000413(
    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{nm000413,
  title = {DRE-LZSH: SEEG benchmark dataset for drug-resistant epilepsy (ictal and interictal segments, 29 patients)},
  author = {Wang Yalin},
  doi = {10.82901/nemar.nm000413},
  url = {https://doi.org/10.82901/nemar.nm000413},
}
§ 02Study · The README

About This Dataset#

Stereo-EEG (SEEG) segments from 29 patients with drug-resistant epilepsy (P001-P029). For each patient there are

seizure (ictal) segments and interictal segments, plus clinical information and channel-level epileptogenic-zone (EZ) annotations. Re-packaged in iEEG-BIDS from the authors’ public release (Zenodo record 17272689, “DRE-LZSH: A SEEG Benchmark Dataset and Pilot Exploration for Advancing Precision Medicine in Drug-Resistant Epilepsy”, CC-BY-4.0). The record describes itself as “the public dataset, running code and results for our paper”. No article was found at conversion time (2026-10-08), and the record does not name the hospital.

participants.tsv reproduces the release table information list/clinical information.xlsx: sex, age at surgery, age

at onset, MRI code (P/N), number of seizures, electrodes and contacts, EZ, therapy and Engel outcome (all patients Engel I). It also includes the channel numbers of information list/channel level EZ annotations.xlsx (EZ channels and deleted, or bad, channels) verbatim. The release does not say whether these numbers index the channel order of the .mat files or the original montage. Some deleted-channel numbers exceed the number of channels in a file, so they are not mapped onto channels.tsv. Note on P014 (added 2026-10-09). In the release file information list/channel level EZ annotations.xlsx, the row of P014 (EZ “Right occipital pole, medial posterior cingulate gyrus, and mid-posterior hippocampus”) has exactly the same EZ channel list (11 23 24 18 19 21 35 ... 112-118) and deleted-channel list (5 6 20 38 121 122) as P007 (EZ “Left inferior frontal gyrus, posterior middle frontal gyrus, and part of the orbitofrontal cortex”). Nine of these numbers (112-118, 121, 122) are larger than the channel count of every P014 file (85 or 107 channels; the clinical table gives 86 contacts). The P014 row is therefore probably a copy of the P007 row in the release. participants.tsv reproduces the release unchanged (checked against the deposit on 2026-10-09); do not use the P014 ez_channel_ids and deleted_channel_ids without confirmation from the authors.

DOI

DRE-LZSH: SEEG benchmark dataset for drug-resistant epilepsy

Recordings

  • task-ictal: release folder ictal state (seizure segments, typically 40-60 s). task-interictal: folder interictal state (longer segments). run-<k> follows the natural order of the release file names (sz1, sz2, …, sz4-1). sub-<P>/sub-<P>_scans.tsv gives the source file of every run. The release does not document where an ictal segment starts relative to seizure onset, so there are no events.

View full README

DOI

DRE-LZSH: SEEG benchmark dataset for drug-resistant epilepsy

Recordings

  • task-ictal: release folder ictal state (seizure segments, typically 40-60 s). task-interictal: folder interictal state (longer segments). run-<k> follows the natural order of the release file names (sz1, sz2, …, sz4-1). sub-<P>/sub-<P>_scans.tsv gives the source file of every run. The release does not document where an ictal segment starts relative to seizure onset, so there are no events.

  • 2000 Hz. Channel names are kept verbatim from the export (POL A'1, EEG C3-Ref, POL ECG, …). Types are inferred from the names (SEEG, EEG for the scalp channels, ECG, MISC; see task-*_channels.json). Units are not stated (n/a); the magnitudes are consistent with volts.

  • The release stores three MATLAB layouts: eeg_data with an MNE-style info struct, data with ch_names, and data with channel_names and a channel count. All three are read as stored. Each segment became one BrainVision run (IEEE float32: the stored float64 values rounded to float32), and the round-trip check compares every run with its source. No filtering, resampling, re-referencing or channel removal was applied.

  • There are no electrode coordinates in the release. electrodes.tsv lists the SEEG contact names with n/a coordinates.

Other release content

  • derivatives/dre-lzsh-results/: the authors’ result tables (Results/: HFO rates and 1/f power-spectrum fits per segment), copied verbatim.

  • code/dre-lzsh-release/: the authors’ code (HFODetector.m and related scripts, a notebook), copied verbatim.

  • sourcedata/zenodo-17272689/DRE-LZSH/: the original .mat files and the two information-list spreadsheets. The only change is that the MATLAB header text “Created on” (file export date, April and June 2025) has its weekday and day masked. The provenance JSON gives the SHA-256 of each file before and after masking.

License

CC-BY-4.0

NEMAR Metadata#

[![DOI](https://img.shields.io/badge/DOI-10.82901%2Fnemar.nm000413-blue)](https://doi.org/10.82901/nemar.nm000413) # DRE-LZSH: SEEG benchmark dataset for drug-resistant epilepsy Stereo-EEG (SEEG) segments from 29 patients with drug-resistant epilepsy (P001-P029). For each patient there are seizure (ictal) segments and interictal segments, plus clinical information and channel-level epileptogenic-zone (EZ) annotations. Re-packaged in iEEG-BIDS from the authors’ public release (Zenodo record [17272689](https://doi.org/10.5281/zenodo.17272689), “DRE-LZSH: A SEEG Benchmark Dataset and Pilot Exploration for Advancing Precision Medicine in Drug-Resistant Epilepsy”, CC-BY-4.0). The record describes itself as “the public dataset, running code and results for our paper”. No article was found at conversion time (2026-10-08), and the record does not name the hospital. ## Participants participants.tsv reproduces the release table information list/clinical information.xlsx: sex, age at surgery, age at onset, MRI code (P/N), number of seizures, electrodes and contacts, EZ, therapy and Engel outcome (all patients Engel I). It also includes the channel numbers of information list/channel level EZ annotations.xlsx (EZ channels and deleted, or bad, channels) verbatim. The release does not say whether these numbers index the channel order of the .mat files or the original montage. Some deleted-channel numbers exceed the number of channels in a file, so they are not mapped onto channels.tsv. Note on P014 (added 2026-10-09). In the release file information list/channel level EZ annotations.xlsx, the row of P014 (EZ “Right occipital pole, medial posterior cingulate gyrus, and mid-posterior hippocampus”) has exactly the same EZ channel list (11 23 24 18 19 21 35 … 112-118) and deleted-channel list (5 6 20 38 121 122) as P007 (EZ “Left inferior frontal gyrus, posterior middle frontal gyrus, and part of the orbitofrontal cortex”). Nine of these numbers (112-118, 121, 122) are larger than the channel count of every P014 file (85 or 107 channels; the clinical table gives 86 contacts). The P014 row is therefore probably a copy of the P007 row in the release. participants.tsv reproduces the release unchanged (checked against the deposit on 2026-10-09); do not use the P014 ez_channel_ids and deleted_channel_ids without confirmation from the authors. ## Recordings - task-ictal: release folder ictal state (seizure segments, typically 40-60 s). task-interictal: folder

interictal state (longer segments). run-<k> follows the natural order of the release file names (sz1, sz2, …, sz4-1). sub-<P>/sub-<P>_scans.tsv gives the source file of every run. The release does not document where an ictal segment starts relative to seizure onset, so there are no events.

  • 2000 Hz. Channel names are kept verbatim from the export (POL A’1, EEG C3-Ref, POL ECG, …). Types are inferred from the names (SEEG, EEG for the scalp channels, ECG, MISC; see task-*_channels.json). Units are not stated (n/a); the magnitudes are consistent with volts.

  • The release stores three MATLAB layouts: eeg_data with an MNE-style info struct, data with ch_names, and data with channel_names and a channel count. All three are read as stored. Each segment became one BrainVision run (IEEE float32: the stored float64 values rounded to float32), and the round-trip check compares every run with its source. No filtering, resampling, re-referencing or channel removal was applied.

  • There are no electrode coordinates in the release. electrodes.tsv lists the SEEG contact names with n/a coordinates.

## Other release content - derivatives/dre-lzsh-results/: the authors’ result tables (Results/: HFO rates and 1/f power-spectrum fits per

segment), copied verbatim.

  • code/dre-lzsh-release/: the authors’ code (HFODetector.m and related scripts, a notebook), copied verbatim.

  • sourcedata/zenodo-17272689/DRE-LZSH/: the original .mat files and the two information-list spreadsheets. The only change is that the MATLAB header text “Created on” (file export date, April and June 2025) has its weekday and day masked. The provenance JSON gives the SHA-256 of each file before and after masking.

## License CC-BY-4.0

License: CC-BY-4.0

Authors:

  • Wang Yalin

Versions:

Version

DOI

Released

current

10.82901/nemar.nm000413
§ 03Cohort · Participants

Cohort#

Dataset Statistics#

Age distribution by gender (n=29, range 12–37 yr, mean 24.7 yr)

101520253035
Female · 10Male · 19

Sex composition

29
subjects
Female
10
Male
19
F : M ratio
0.53 : 1
34% female · n = 29 subjects with reported sex.

Channel counts (ch)

7379838586919395104107108111113114117118121123124126128139149163

Sampling frequencies (Hz)

2000.020002000.0

Total recording duration: 7 h 36 min

§ 04Signal · Electrodes & trace

Signal · Electrodes & live trace#

Fig. 01 Signal & montage 107 (18), 113 (14), 91 (14), 117 (13), 111 (10), 123 (10), 163 (10), 93 (10), 126 (9), 79 (8), 108 (8), 83 (7), 95 (6), 118 (6), 128 (6), 139 (6), 114 (6), 85 (6), 124 (6), 73 (4), 149 (4), 121 (4), 86 (3), 104 (2) ch · iEEG · 2000 Hz · 29 subjects, 190 recordings
Live trace viewer — sub-P001 · task-ictal · run-05

Showing one representative recording out of 29 subjects and 190 recordings in this dataset. Browse the full set on OpenNeuro; drop any other _ieeg.{set,edf,bdf,vhdr} file onto the viewer (or pass ?ieeg=<url>) to inspect it.

No scalp electrode layout is currently indexed for this dataset. Once the eegdash montage registry ingests it, the interactive viewer will appear here automatically.

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.

NEMAR pipeline statistics — NM000413
HED event descriptors word cloud HED event descriptors word cloud — NM000413
§ 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…
Full dataset metadata table

Dataset ID

NM000413

Title

DRE-LZSH: SEEG benchmark dataset for drug-resistant epilepsy (ictal and interictal segments, 29 patients)

Author (year)

—

Canonical

—

Importable as

NM000413

Year

20

Authors

Wang Yalin

License

CC-BY-4.0

Citation / DOI

10.82901/nemar.nm000413

Source links

OpenNeuro | NeMAR | Source URL
Copy-paste BibTeX
@dataset{nm000413,
  title = {DRE-LZSH: SEEG benchmark dataset for drug-resistant epilepsy (ictal and interictal segments, 29 patients)},
  author = {Wang Yalin},
  doi = {10.82901/nemar.nm000413},
  url = {https://doi.org/10.82901/nemar.nm000413},
}
§ 06API · Programmatic access

API Reference#

Signature
eegdash.dataset
class
eegdash.dataset.NM000413(cache_dir, query=None, s3_bucket=None, **kwargs)
Bases: EEGDashDataset
Author (year)—
Canonical—
Importable asNM000413
Sourceeegdash/dataset/registry.py · [source ↗]
class eegdash.dataset.NM000413(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#

DRE-LZSH: SEEG benchmark dataset for drug-resistant epilepsy (ictal and interictal segments, 29 patients)

Study:

nm000413 (NeMAR)

Author (year):

—

Canonical:

—

Also importable as: NM000413.

Modality: ieeg; Subject type: Unknown. Subjects: 29; recordings: 190; tasks: 2.

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/nm000413 NeMAR dataset: https://nemar.org/dataexplorer/detail?dataset_id=nm000413 DOI: https://doi.org/10.82901/nemar.nm000413

Examples

>>> from eegdash.dataset import NM000413
>>> dataset = NM000413(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 — NM000413.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 nm000413 to reproduce the tutorial on this dataset.

Citation

Wang Yalin (20). DRE-LZSH: SEEG benchmark dataset for drug-resistant epilepsy (ictal and interictal segments, 29 patients). 10.82901/nemar.nm000413

Provenance

¹Contributed to nemar in BIDS format.

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

³Persistent identifier: 10.82901/nemar.nm000413.

BIDS
BIDS 1.10.0
Sidecars
channels · eeg.json
Provenance
CC-BY-4.0 · 10.82901/nemar.nm000413
Machine-readable
schema.org/Dataset · Croissant
Mirrors
OpenNeuro · NEMAR · HF org

See Also#