EEGdash›NeMAR›NM000364
Iss. 364 · 7 subjects · 8 recordings · CC-BY-NC-4.0
Dataset Brief · Task-SEEG recordings from epilepsy patients performing a visu…

NM000364: ieeg dataset, 7 subjects#

Task-SEEG recordings from epilepsy patients performing a visual working memory task (Tian, Figshare 32189838)

Access recordings and metadata through EEGDash.

Citation: Ziwei Tian (2026). Task-SEEG recordings from epilepsy patients performing a visual working memory task (Tian, Figshare 32189838). 10.82901/nemar.nm000364

Modality: ieeg Subjects: 7 Recordings: 8 License: CC-BY-NC-4.0 Source: nemar

Metadata: Complete (100%)

7-participant iEEG dataset — Task-SEEG recordings from epilepsy patients performing a visual working memory task (Tian, Figshare 32189838).

Data & curation Ziwei Tian
Senior author Ziwei Tian
Year 2026 · Distributed via NeMAR
iEEG · 148 ch2048 HzBIDS 1.10.0Task · vwm
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 NM000364

dataset = NM000364(cache_dir="./data")
# Get the raw object of the first recording
raw = dataset.datasets[0].raw
print(raw.info)

Filter by subject

dataset = NM000364(cache_dir="./data", subject="01")

Advanced query

dataset = NM000364(
    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{nm000364,
  title = {Task-SEEG recordings from epilepsy patients performing a visual working memory task (Tian, Figshare 32189838)},
  author = {Ziwei Tian},
  doi = {10.82901/nemar.nm000364},
  url = {https://doi.org/10.82901/nemar.nm000364},
}
§ 02Study · The README

About This Dataset#

This is a BIDS (iEEG) copy of the Figshare dataset

Tian, Ziwei (2026). task-SEEG data from epilepsy patients. figshare. Dataset. https://doi.org/10.6084/m9.figshare.32189838.v1 (article 32189838, version 1, published 2026-05-07)

DOI

Task-SEEG recordings from epilepsy patients performing a visual working memory task

Source description (Figshare, verbatim): “This dataset contains stereo-electroencephalography (SEEG) recordings acquired from epilepsy patients while they performed a visual working memory task. The data were collected to investigate the neural mechanisms underlying working memory processes in the human brain.” Participants: “patients with drug-resistant epilepsy who underwent SEEG implantation for clinical evaluation. Electrode placement was determined solely by clinical requirements, with no additional electrodes implanted for research purposes.” Electrodes: “Intracranial multi-contact depth electrodes (8–16 contacts; length, 2 mm; diameter, 0.8 mm; spacing,

View full README

DOI

Task-SEEG recordings from epilepsy patients performing a visual working memory task

Source description (Figshare, verbatim): “This dataset contains stereo-electroencephalography (SEEG) recordings acquired from epilepsy patients while they performed a visual working memory task. The data were collected to investigate the neural mechanisms underlying working memory processes in the human brain.” Participants: “patients with drug-resistant epilepsy who underwent SEEG implantation for clinical evaluation. Electrode placement was determined solely by clinical requirements, with no additional electrodes implanted for research purposes.” Electrodes: “Intracranial multi-contact depth electrodes (8–16 contacts; length, 2 mm; diameter, 0.8 mm; spacing, 1.5 mm; Huake-Hengsheng Medical Technology, Beijing, China) were stereotactically implanted with robotic assistance.” Recording: “Amplifier: Nicolet EEG; Sampling rate: 2048 Hz; Online reference: A white matter site selected by the clinical team.” Task: “Participants performed a visual working memory task. Further details of the experimental paradigm can be found in the accompanying research article.”

Contents

  • 7 participants (sub-01 to sub-07), 8 continuous EDF recordings (sub-03 has two runs), task label vwm.

  • 2048 Hz, 148 signals per recording (the EDF+ annotation signal of sub-06 and sub-07 is not counted).

  • Total duration 22,168 s (6.2 h): 1,080.6 to 4,010.9 s per recording (exact values in each _ieeg.json).

  • Units as in the EDF headers (µV for SEEG contacts). No filtering, resampling, re-referencing or channel removal was done for this copy. The EDF prefilter fields are empty, so low_cutoff/high_cutoff are n/a.

  • participants.tsv: age, sex, years of education, seizure-onset zone, and the contacts the authors list as frontal eye field and hippocampus. Values are verbatim from the source (columns renamed to snake_case; male/female coded M/F).

What the source does not contain

  • No events, trial or behavioural files. The Figshare record has none, and the EDFs have no task annotations (sub-07 has one EDF+ annotation “renwu”, pinyin for “task”, at 0.75 s; sub-06 and sub-07 also have machine log entries). The TRIG channel carries a few analog-like transitions per recording (about 80 to 170 sample-to-sample changes), not a clean digital code. It was kept but not decoded into events. Users need the paradigm from the authors’ article.

  • No electrode coordinates and no imaging. electrodes.tsv lists the SEEG contact names with x/y/z = n/a. coordsystem.json says “Other” with units n/a.

  • No ethics statement and no funding statement are given in the record.

Channel types (changed from the source)

The source channels.tsv (MNE-BIDS output) typed all 148 signals as SEEG. Here the types follow the EDF labels:

SEEG for electrode contacts (letter, optional prime marks, number); ADC for DC1–DC16 (Nicolet DC auxiliary inputs); TRIG for TRIG; MISC for OSAT, PR, Pleth and for amplifier inputs with the Nicolet default names C123–C128, which the source does not assign to an electrode. Channels whose digital value is constant over the whole recording are marked bad with status_description “flat”. The source channel tables are kept in sourcedata/.

Power-line frequency

The source sidecars say “n/a”. PowerLineFrequency is set to 50 because the recordings show a clear 50 Hz peak (spectral peak-to-neighbour ratio 18 to 6,400 at 50 Hz, about 1 at 60 Hz, in every recording).

Licence

CC-BY-NC-4.0

Associated publications

The Figshare record refers to “the accompanying research article” but does not name it. These articles by the same first author describe SEEG recordings during a visual working memory task with eye tracking in epilepsy patients. The link is inferred from author, paradigm and dates; the record does not state it: - Tian Z, Huang S, Liu D, Yang Z, Li S, Hu B, Feng L, Wang Q. Hippocampal interictal spikes disrupt theta-band

hippocampal-frontal eye field connectivity: fixation skewness as an indicator of transient network instability. Epilepsy & Behavior 181:111071 (2026). https://doi.org/10.1016/j.yebeh.2026.111071 (seven epilepsy patients with hippocampal and frontal-eye-field iEEG; published two days after the Figshare record).

  • Tian Z, Huang S, Liu D, Yang Z, Li S, Hu B, Wang Q, Feng L. Distinct and coordinated contributions of hippocampus and frontal eye field to novelty exploration and revisitation. NeuroImage 329:121838 (2026). https://doi.org/10.1016/j.neuroimage.2026.121838

Both list affiliations at the Xi’an Institute of Optics and Precision Mechanics (CAS) and Xiangya Hospital, Central South University. The Figshare record names neither institution.

De-identification done for this copy

The source EDF headers contain a hospital record number, the date of birth and the real recording date. The source scans.tsv files contain the real recording dates. The EDF+ annotation signals of sub-06 and sub-07 hold a Montage: entry that contains a personal name (the files do not say whose; it may be the name of a montage or of a staff member, or a patient). For this copy:

Dates truncated to month (day set to 01). - EDF patient field set to the EDF+ form X <sex> X X: sex (M/F, same as participants.tsv) is kept; the hospital

record number and the date of birth are removed (birthdate subfield X; EDF+ allows only a full date or X). The year of birth from the source header is kept in participants.tsv (birth_year; day and month removed); it agrees within one year with age and the recording year for every participant. Recording field set to Startdate 01-MMM-YYYY X X X and start date to 01.MM.YY, keeping the source year and month. The time of day is kept.

  • In sub-06 and sub-07, the text after Montage: in the EDF+ annotation signal is replaced by X characters of the same byte length. Other annotations (“Clip Note”, “Gain/Filter Change”, “renwu”) are kept.

  • scans.tsv acq_time keeps year, month and time of day, with the day set to 01. The two sub-03 runs were recorded on the same day, so the interval between them is kept.

  • A byte comparison against the source confirmed that no other byte changed: all signal samples are identical to the Figshare files.

  • The source EDF and scans.tsv files are therefore not included in sourcedata/. The other source files (README, dataset_description, participants, channels and sidecar JSON) are included unchanged in sourcedata/figshare-32189838-v1/. PROVENANCE.tsv there lists every source file with its Figshare MD5 and whether it is included.

NEMAR Metadata#

[![DOI](https://img.shields.io/badge/DOI-10.82901%2Fnemar.nm000364-blue)](https://doi.org/10.82901/nemar.nm000364) # Task-SEEG recordings from epilepsy patients performing a visual working memory task This is a BIDS (iEEG) copy of the Figshare dataset > Tian, Ziwei (2026). task-SEEG data from epilepsy patients. figshare. Dataset. > https://doi.org/10.6084/m9.figshare.32189838.v1 (article 32189838, version 1, published 2026-05-07) Source description (Figshare, verbatim): “This dataset contains stereo-electroencephalography (SEEG) recordings acquired from epilepsy patients while they performed a visual working memory task. The data were collected to investigate the neural mechanisms underlying working memory processes in the human brain.” Participants: “patients with drug-resistant epilepsy who underwent SEEG implantation for clinical evaluation. Electrode placement was determined solely by clinical requirements, with no additional electrodes implanted for research purposes.” Electrodes: “Intracranial multi-contact depth electrodes (8–16 contacts; length, 2 mm; diameter, 0.8 mm; spacing, 1.5 mm; Huake-Hengsheng Medical Technology, Beijing, China) were stereotactically implanted with robotic assistance.” Recording: “Amplifier: Nicolet EEG; Sampling rate: 2048 Hz; Online reference: A white matter site selected by the clinical team.” Task: “Participants performed a visual working memory task. Further details of the experimental paradigm can be found in the accompanying research article.” ## Contents - 7 participants (sub-01 to sub-07), 8 continuous EDF recordings (sub-03 has two runs), task label vwm. - 2048 Hz, 148 signals per recording (the EDF+ annotation signal of sub-06 and sub-07 is not counted). - Total duration 22,168 s (6.2 h): 1,080.6 to 4,010.9 s per recording (exact values in each _ieeg.json). - Units as in the EDF headers (µV for SEEG contacts). No filtering, resampling, re-referencing or channel removal

was done for this copy. The EDF prefilter fields are empty, so low_cutoff/high_cutoff are n/a.

  • participants.tsv: age, sex, years of education, seizure-onset zone, and the contacts the authors list as frontal eye field and hippocampus. Values are verbatim from the source (columns renamed to snake_case; male/female coded M/F).

## What the source does not contain - No events, trial or behavioural files. The Figshare record has none, and the EDFs have no task annotations

(sub-07 has one EDF+ annotation “renwu”, pinyin for “task”, at 0.75 s; sub-06 and sub-07 also have machine log entries). The TRIG channel carries a few analog-like transitions per recording (about 80 to 170 sample-to-sample changes), not a clean digital code. It was kept but not decoded into events. Users need the paradigm from the authors’ article.

  • No electrode coordinates and no imaging. electrodes.tsv lists the SEEG contact names with x/y/z = n/a. coordsystem.json says “Other” with units n/a.

  • No ethics statement and no funding statement are given in the record.

## Channel types (changed from the source) The source channels.tsv (MNE-BIDS output) typed all 148 signals as SEEG. Here the types follow the EDF labels: SEEG for electrode contacts (letter, optional prime marks, number); ADC for DC1–DC16 (Nicolet DC auxiliary inputs); TRIG for TRIG; MISC for OSAT, PR, Pleth and for amplifier inputs with the Nicolet default names C123–C128, which the source does not assign to an electrode. Channels whose digital value is constant over the whole recording are marked bad with status_description “flat”. The source channel tables are kept in sourcedata/. ## Power-line frequency The source sidecars say “n/a”. PowerLineFrequency is set to 50 because the recordings show a clear 50 Hz peak (spectral peak-to-neighbour ratio 18 to 6,400 at 50 Hz, about 1 at 60 Hz, in every recording). ## Licence CC-BY-NC-4.0 ## Associated publications The Figshare record refers to “the accompanying research article” but does not name it. These articles by the same first author describe SEEG recordings during a visual working memory task with eye tracking in epilepsy patients. The link is inferred from author, paradigm and dates; the record does not state it: - Tian Z, Huang S, Liu D, Yang Z, Li S, Hu B, Feng L, Wang Q. Hippocampal interictal spikes disrupt theta-band

hippocampal-frontal eye field connectivity: fixation skewness as an indicator of transient network instability. Epilepsy & Behavior 181:111071 (2026). https://doi.org/10.1016/j.yebeh.2026.111071 (seven epilepsy patients with hippocampal and frontal-eye-field iEEG; published two days after the Figshare record).

  • Tian Z, Huang S, Liu D, Yang Z, Li S, Hu B, Wang Q, Feng L. Distinct and coordinated contributions of hippocampus and frontal eye field to novelty exploration and revisitation. NeuroImage 329:121838 (2026). https://doi.org/10.1016/j.neuroimage.2026.121838

Both list affiliations at the Xi’an Institute of Optics and Precision Mechanics (CAS) and Xiangya Hospital, Central South University. The Figshare record names neither institution. ## De-identification done for this copy The source EDF headers contain a hospital record number, the date of birth and the real recording date. The source scans.tsv files contain the real recording dates. The EDF+ annotation signals of sub-06 and sub-07 hold a Montage: entry that contains a personal name (the files do not say whose; it may be the name of a montage or of a staff member, or a patient). For this copy: Dates truncated to month (day set to 01). - EDF patient field set to the EDF+ form X <sex> X X: sex (M/F, same as participants.tsv) is kept; the hospital

record number and the date of birth are removed (birthdate subfield X; EDF+ allows only a full date or X). The year of birth from the source header is kept in participants.tsv (birth_year; day and month removed); it agrees within one year with age and the recording year for every participant. Recording field set to Startdate 01-MMM-YYYY X X X and start date to 01.MM.YY, keeping the source year and month. The time of day is kept.

  • In sub-06 and sub-07, the text after Montage: in the EDF+ annotation signal is replaced by X characters of the same byte length. Other annotations (“Clip Note”, “Gain/Filter Change”, “renwu”) are kept.

  • scans.tsv acq_time keeps year, month and time of day, with the day set to 01. The two sub-03 runs were recorded on the same day, so the interval between them is kept.

  • A byte comparison against the source confirmed that no other byte changed: all signal samples are identical to the Figshare files.

  • The source EDF and scans.tsv files are therefore not included in sourcedata/. The other source files (README, dataset_description, participants, channels and sidecar JSON) are included unchanged in sourcedata/figshare-32189838-v1/. PROVENANCE.tsv there lists every source file with its Figshare MD5 and whether it is included.

License: CC-BY-NC-4.0

Authors:

  • Ziwei Tian

Versions:

Version

DOI

Released

current

10.82901/nemar.nm000364
§ 03Cohort · Participants

Cohort#

Dataset Statistics#

Age distribution by gender (n=7, range 18–56 yr, mean 29.0 yr)

1520253055
Female · 1Male · 6

Sex composition

7
subjects
Female
1
Male
6
F : M ratio
0.17 : 1
14% female · n = 7 subjects with reported sex.

Channel counts: 148 ch (n=8 recordings)

Sampling frequencies: 2048.0 Hz (n=8 recordings)

Total recording duration: 6 h 9 min

§ 04Signal · Electrodes & trace

Signal · Electrodes & live trace#

Fig. 01 Signal & montage 148 ch · iEEG · 2048 Hz · 7 subjects, 8 recordings
Live trace viewer — sub-04 · task-vwm · run-01

Showing one representative recording out of 7 subjects and 8 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 — NM000364
HED event descriptors word cloud HED event descriptors word cloud — NM000364
§ 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

NM000364

Title

Task-SEEG recordings from epilepsy patients performing a visual working memory task (Tian, Figshare 32189838)

Author (year)

—

Canonical

—

Importable as

NM000364

Year

2026

Authors

Ziwei Tian

License

CC-BY-NC-4.0

Citation / DOI

10.82901/nemar.nm000364

Source links

OpenNeuro | NeMAR | Source URL
Copy-paste BibTeX
@dataset{nm000364,
  title = {Task-SEEG recordings from epilepsy patients performing a visual working memory task (Tian, Figshare 32189838)},
  author = {Ziwei Tian},
  doi = {10.82901/nemar.nm000364},
  url = {https://doi.org/10.82901/nemar.nm000364},
}
§ 06API · Programmatic access

API Reference#

Signature
eegdash.dataset
class
eegdash.dataset.NM000364(cache_dir, query=None, s3_bucket=None, **kwargs)
Bases: EEGDashDataset
Author (year)—
Canonical—
Importable asNM000364
Sourceeegdash/dataset/registry.py · [source ↗]
class eegdash.dataset.NM000364(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#

Task-SEEG recordings from epilepsy patients performing a visual working memory task (Tian, Figshare 32189838)

Study:

nm000364 (NeMAR)

Author (year):

—

Canonical:

—

Also importable as: NM000364.

Modality: ieeg; Subject type: Unknown. Subjects: 7; recordings: 8; 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/nm000364 NeMAR dataset: https://nemar.org/dataexplorer/detail?dataset_id=nm000364 DOI: https://doi.org/10.82901/nemar.nm000364

Examples

>>> from eegdash.dataset import NM000364
>>> dataset = NM000364(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 — NM000364.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 nm000364 to reproduce the tutorial on this dataset.

Citation

Ziwei Tian (2026). Task-SEEG recordings from epilepsy patients performing a visual working memory task (Tian, Figshare 32189838). 10.82901/nemar.nm000364

Provenance

¹Contributed to nemar in BIDS format.

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

³Persistent identifier: 10.82901/nemar.nm000364.

BIDS
BIDS 1.10.0
Sidecars
channels · electrodes · coordsystem · eeg.json
Provenance
CC-BY-NC-4.0 · 10.82901/nemar.nm000364
Machine-readable
schema.org/Dataset · Croissant
Mirrors
OpenNeuro · NEMAR · HF org

See Also#