EEGdash›NeMAR›NM000462

Iss. 462 · 3 subjects · 4 recordings · CC-BY-NC-SA-4.0

Dataset Brief · stim-seizures (Ojemann et al.; HUP + CHOP)

NM000462: ieeg dataset, 3 subjects#

stim-seizures (Ojemann et al.; HUP + CHOP): seizure annotations, electrode localisations and preprocessed SEEG seizure traces (derivative)

Access recordings and metadata through EEGDash.

Citation: William K. Ojemann, Brian Litt, Erin Conrad (20). stim-seizures (Ojemann et al.; HUP + CHOP): seizure annotations, electrode localisations and preprocessed SEEG seizure traces (derivative). 10.82901/nemar.nm000462

Modality: ieeg Subjects: 3 Recordings: 4 License: CC-BY-NC-SA-4.0 Source: nemar

Metadata: Complete (100%)

3-participant iEEG dataset — stim-seizures (Ojemann et al.; HUP + CHOP): seizure annotations, electrode localisations and preprocessed SEEG seizure traces (derivative).

Data & curation William K. Ojemann · Brian Litt · Erin Conrad
Senior author Erin Conrad
Year 20 · Distributed via NeMAR

iEEG · 70, 69, 149, 153 ch128 HzBIDS 1.10.0Task · ictal

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#

Get Started

Install

pip install eegdash

Access the data

from eegdash.dataset import NM000462

dataset = NM000462(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 = NM000462(cache_dir="./data", subject="01")

Advanced query

dataset = NM000462(
    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{nm000462,
  title = {stim-seizures (Ojemann et al.; HUP + CHOP): seizure annotations, electrode localisations and preprocessed SEEG seizure traces (derivative)},
  author = {William K. Ojemann and Brian Litt and Erin Conrad},
  doi = {10.82901/nemar.nm000462},
  url = {https://doi.org/10.82901/nemar.nm000462},
}

§ 02Study · The README

About This Dataset#

This is a derivative dataset. It repackages the processed data released by Ojemann, Litt and Conrad on Pennsieve Discover

(“stim-seizures”, dataset 455 version 1, doi:10.26275/toip-zio0, Penn CNT) for the study “Can electrical stimulation replace spontaneous seizures?” (preprint: Ojemann et al., medRxiv 2025, doi:10.1101/2025.08.29.25334082; later version titled “Automated annotation of low-frequency stimulation-induced seizures uncovers seizure generating networks”).

The release contains metadata and processed data checkpoints for the manuscript’s analyses and figures. It contains **no raw iEEG

recordings**. Code: penn-cnt/stim-seizures-manuscript Patients were evaluated with intracranial EEG at the Hospital of the University of Pennsylvania (HUP) and the Children’s Hospital of Philadelphia (CHOP) and received low-frequency (1 Hz) electrical stimulation. Related NEMAR dataset: nm000402 (Multi-Expert Seizure Annotation, Penn CNT / CHOP, shared authors). Raw HUP recordings of this group are hosted on iEEG.org (recording names in the recording column, e.g. HUP224_phaseII).

DOI

stim-seizures (Pennsieve 455): processed data, derivative BIDS-iEEG

Contents

participants.tsv: all 105 patients of the release (55 HUP, 50 CHOP). Columns come from the released METADATA/stim_seizure_information - metadata-4.csv (HUP) and METADATA/CHOP_metadata.csv (CHOP), copied as released (see participants.json). HUP sex codes 1/2 are mapped to M/F as in the authors’ code/metadata_notebook.ipynb. Age at recording is released only for CHOP patients (rounded here to 0.1 y). 72 of the listed patients appear only

View full README

DOI

stim-seizures (Pennsieve 455): processed data, derivative BIDS-iEEG

Contents

participants.tsv: all 105 patients of the release (55 HUP, 50 CHOP). Columns come from the released METADATA/stim_seizure_information - metadata-4.csv (HUP) and METADATA/CHOP_metadata.csv (CHOP), copied as released (see participants.json). HUP sex codes 1/2 are mapped to M/F as in the authors’ code/metadata_notebook.ipynb. Age at recording is released only for CHOP patients (rounded here to 0.1 y). 72 of the listed patients appear only in the metadata tables (no seizure rows and no processed files), so they have no subject folder.
sub-<label>/ieeg/sub-<label>_seizures.tsv (33 patients; listed in .bidsignore, not a BIDS file type): one row per seizure from two released tables: stim_seizure_information_BIDS.csv (clinician-marked times: approximate_onset, offset (= end), EEC, UEO in seconds from the start of the iEEG.org recording; stim flag, stimulation channels, typical semiology, mtl) and the optimized LSTM model annotations (optimized_predicted_channels_LSTM_tuned_thresholds_v4_sz-mean_pt-mean_smooth-med.pkl: model, threshold, onset (ueo) and spread (sec) channel sets, strict/loose). Values are copied; lists are space-separated.
sub-<label>/ieeg/sub-<label>_electrodes.tsv + _coordsystem.json (31 patients): contact coordinates and atlas labels from PROCESSED_DATA/<patient>/electrode_localizations_dkt.pkl (HUP) or electrode_localizations_CHOPR.pkl (CHOP). The release does not state the coordinate space or units, so the coordinate system is Other and units n/a.
sub-HUP224, sub-HUP235, sub-HUP238 _task-ictal_acq-<spontaneous|stiminduced>_ieeg.vhdr: the four seizure traces the authors released for figure plotting (plot_sz_224/235/238.pkl, plot_st_sz_224.pkl): preprocessed bipolar SEEG, 128 Hz (the value used in the authors’ plotting notebook), HUP224 spontaneous 331.0 s x 153 channels; HUP224 stiminduced 275.0 s x 149 channels; HUP235 spontaneous 371.0 s x 70 channels; HUP238 spontaneous 492.0 s x 69 channels. Amplitude units are not stated in the release (units n/a). Stored as BrainVision IEEE float32; round trip checked (float32 values identical to the released float64 values cast to float32).
sourcedata/pennsieve-455-v1/: every released file (pickled DataFrames, HDF5 spread matrices, stim_info*.mat, metadata CSVs, figure PDFs, Pennsieve manifest/readme). SOURCE_MANIFEST.json lists the original SHA-256 of each file. Files are byte-identical to the release except file-creation timestamps in MAT headers and PDF metadata, which were set to day 01 (time removed, same length).

Privacy

All files were scanned (text, CSV, MAT strings, HDF5, pickled DataFrames, PDF metadata) for names, initials, record numbers, full dates and ages above 89. Patient codes (HUP###, CHOP###) are the pseudonyms used in the public release. Seizure times are seconds from recording start. No ages above 89. Key status: destroyed (we hold no mapping to identities).

Licence

CC-BY-NC-SA-4.0, as on the source record (Pennsieve: “Creative Commons Attribution - NonCommercial-ShareAlike”; DataCite rightsList:

Creative Commons Attribution Non Commercial Share Alike 4.0 International, cc-by-nc-sa-4.0).

Ethics

Verbatim from the preprint (medRxiv doi:10.1101/2025.08.29.25334082, Author Declarations): “The Institutional Review Boards (IRBs) of the Hospital of the University of Pennsylvania (HUP) and the Children’s Hospital of Philadelphia (CHOP) approved retrospective analysis of EEG and clinical data for all patients. Stimulation was performed under IRB-approved research by the IRB at each institution (14 HUP patients; all CHOP patients) or clinical protocols (HUP, N = 41).”

NEMAR Metadata#

[![DOI](https://img.shields.io/badge/DOI-10.82901%2Fnemar.nm000462-blue)](https://doi.org/10.82901/nemar.nm000462) # stim-seizures (Pennsieve 455): processed data, derivative BIDS-iEEG This is a derivative dataset. It repackages the processed data released by Ojemann, Litt and Conrad on Pennsieve Discover (“stim-seizures”, dataset 455 version 1, doi:10.26275/toip-zio0, Penn CNT) for the study “Can electrical stimulation replace spontaneous seizures?” (preprint: Ojemann et al., medRxiv 2025, doi:10.1101/2025.08.29.25334082; later version titled “Automated annotation of low-frequency stimulation-induced seizures uncovers seizure generating networks”). The release contains metadata and processed data checkpoints for the manuscript’s analyses and figures. It contains no raw iEEG recordings. Code: penn-cnt/stim-seizures-manuscript Patients were evaluated with intracranial EEG at the Hospital of the University of Pennsylvania (HUP) and the Children’s Hospital of Philadelphia (CHOP) and received low-frequency (1 Hz) electrical stimulation. Related NEMAR dataset: nm000402 (Multi-Expert Seizure Annotation, Penn CNT / CHOP, shared authors). Raw HUP recordings of this group are hosted on iEEG.org (recording names in the recording column, e.g. HUP224_phaseII). ## Contents - participants.tsv: all 105 patients of the release (55 HUP, 50 CHOP). Columns

come from the released METADATA/stim_seizure_information - metadata-4.csv (HUP) and METADATA/CHOP_metadata.csv (CHOP), copied as released (see participants.json). HUP sex codes 1/2 are mapped to M/F as in the authors’ code/metadata_notebook.ipynb. Age at recording is released only for CHOP patients (rounded here to 0.1 y). 72 of the listed patients appear only in the metadata tables (no seizure rows and no processed files), so they have no subject folder.

sub-<label>/ieeg/sub-<label>_seizures.tsv (33 patients; listed in .bidsignore, not a BIDS file type): one row per seizure from two released tables: stim_seizure_information_BIDS.csv (clinician-marked times: approximate_onset, offset (= end), EEC, UEO in seconds from the start of the iEEG.org recording; stim flag, stimulation channels, typical semiology, mtl) and the optimized LSTM model annotations (optimized_predicted_channels_LSTM_tuned_thresholds_v4_sz-mean_pt-mean_smooth-med.pkl: model, threshold, onset (ueo) and spread (sec) channel sets, strict/loose). Values are copied; lists are space-separated.
sub-<label>/ieeg/sub-<label>_electrodes.tsv + _coordsystem.json (31 patients): contact coordinates and atlas labels from PROCESSED_DATA/<patient>/electrode_localizations_dkt.pkl (HUP) or electrode_localizations_CHOPR.pkl (CHOP). The release does not state the coordinate space or units, so the coordinate system is Other and units n/a.
sub-HUP224, sub-HUP235, sub-HUP238 _task-ictal_acq-<spontaneous|stiminduced>_ieeg.vhdr: the four seizure traces the authors released for figure plotting (plot_sz_224/235/238.pkl, plot_st_sz_224.pkl): preprocessed bipolar SEEG, 128 Hz (the value used in the authors’ plotting notebook), HUP224 spontaneous 331.0 s x 153 channels; HUP224 stiminduced 275.0 s x 149 channels; HUP235 spontaneous 371.0 s x 70 channels; HUP238 spontaneous 492.0 s x 69 channels. Amplitude units are not stated in the release (units n/a). Stored as BrainVision IEEE float32; round trip checked (float32 values identical to the released float64 values cast to float32).
sourcedata/pennsieve-455-v1/: every released file (pickled DataFrames, HDF5 spread matrices, stim_info*.mat, metadata CSVs, figure PDFs, Pennsieve manifest/readme). SOURCE_MANIFEST.json lists the original SHA-256 of each file. Files are byte-identical to the release except file-creation timestamps in MAT headers and PDF metadata, which were set to day 01 (time removed, same length).

## Privacy All files were scanned (text, CSV, MAT strings, HDF5, pickled DataFrames, PDF metadata) for names, initials, record numbers, full dates and ages above 89. Patient codes (HUP###, CHOP###) are the pseudonyms used in the public release. Seizure times are seconds from recording start. No ages above 89. Key status: destroyed (we hold no mapping to identities). ## Licence CC-BY-NC-SA-4.0, as on the source record (Pennsieve: “Creative Commons Attribution - NonCommercial-ShareAlike”; DataCite rightsList: Creative Commons Attribution Non Commercial Share Alike 4.0 International, cc-by-nc-sa-4.0). ## Ethics Verbatim from the preprint (medRxiv doi:10.1101/2025.08.29.25334082, Author Declarations): “The Institutional Review Boards (IRBs) of the Hospital of the University of Pennsylvania (HUP) and the Children’s Hospital of Philadelphia (CHOP) approved retrospective analysis of EEG and clinical data for all patients. Stimulation was performed under IRB-approved research by the IRB at each institution (14 HUP patients; all CHOP patients) or clinical protocols (HUP, N = 41).”

License: CC-BY-NC-SA-4.0

Authors:

William K. Ojemann
Brian Litt
Erin Conrad

Versions:

Version	DOI	Released
`current`	10.82901/nemar.nm000462

§ 03Cohort · Participants

Cohort#

Dataset Statistics#

Age distribution (n=47, range 1–23 yr, mean 12.6 yr · sex per subject not reported)

05101520

Sex composition

Female
54
Male
48
F : M ratio
1.12 : 1

53% female · n = 102 subjects with reported sex.

HandednessRight · 38Left · 5

Channel counts (ch)

6970149153

Sampling frequencies: 128.0 Hz (n=4 recordings)

Total recording duration: 24 min

§ 04Signal · Electrodes & trace

Signal · Electrodes & live trace#

Fig. 01 Signal & montage 70, 69, 149, 153 ch · iEEG · 128 Hz · 3 subjects, 4 recordings

Live trace viewer — sub-HUP235 · task-ictal

Showing one representative recording out of 3 subjects and 4 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.

Electrode layout — iEEG · 232 sensors — 232 channels

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

§ 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…

§ 06API · Programmatic access

API Reference#

Signature

eegdash.dataset

class

eegdash.dataset.NM000462(cache_dir, query=None, s3_bucket=None, **kwargs)

Bases: EEGDashDataset

Author (year)—

Canonical—

Importable asNM000462

Sourceeegdash/dataset/registry.py · [source ↗]

class eegdash.dataset.NM000462(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#

stim-seizures (Ojemann et al.; HUP + CHOP): seizure annotations, electrode localisations and preprocessed SEEG seizure traces (derivative)

Study:: nm000462 (NeMAR)
Author (year):: —
Canonical:: —

Also importable as: NM000462.

Modality: ieeg; Subject type: Unknown. Subjects: 3; recordings: 4; 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/nm000462 NeMAR dataset: https://nemar.org/dataexplorer/detail?dataset_id=nm000462 DOI: https://doi.org/10.82901/nemar.nm000462

Examples

>>> from eegdash.dataset import NM000462
>>> dataset = NM000462(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 — NM000462.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 nm000462 to reproduce the tutorial on this dataset.

Citation

William K. Ojemann, Brian Litt, Erin Conrad (20). stim-seizures (Ojemann et al.; HUP + CHOP): seizure annotations, electrode localisations and preprocessed SEEG seizure traces (derivative). 10.82901/nemar.nm000462

Provenance

¹Contributed to nemar in BIDS format.

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

³Persistent identifier: 10.82901/nemar.nm000462.

Related & sibling datasets

NM000472IEEG · 1 subj NM000400IEEG · 7 subj NM000471IEEG · 11 subj NM000437IEEG · 5 subj NM000469IEEG · 1 subj

+ 1 more — see See Also below →

BIDS

BIDS 1.10.0

Sidecars

channels · electrodes · coordsystem · eeg.json

Provenance

CC-BY-NC-SA-4.0 · 10.82901/nemar.nm000462

Machine-readable

schema.org/Dataset · Croissant

Mirrors

OpenNeuro · NEMAR · HF org

Dataset ID	`NM000462`
Title	stim-seizures (Ojemann et al.; HUP + CHOP): seizure annotations, electrode localisations and preprocessed SEEG seizure traces (derivative)
Author (year)	—
Canonical	—
Importable as	`NM000462`
Year	20
Authors	William K. Ojemann, Brian Litt, Erin Conrad
License	CC-BY-NC-SA-4.0
Citation / DOI	10.82901/nemar.nm000462
Source links	OpenNeuro \| NeMAR \| Source URL

NM000462: ieeg dataset, 3 subjects#

Quickstart#

About This Dataset#

NEMAR Metadata#

Cohort#

Dataset Statistics#

Signal · Electrodes & live trace#

NEMAR Processing Statistics#

Manifest#

File Explorer#

API Reference#

Citation

Provenance

Related & sibling datasets

See Also#