NM000362: ieeg dataset, 106 subjects#
MNI Open iEEG Atlas: normal intracranial EEG in wakefulness and sleep (processed 200 Hz atlas clips)
Access recordings and metadata through EEGDash.
Citation: Birgit Frauscher, Nicolás von Ellenrieder, Rina Zelmann, Irena Doležalová, Lorella Minotti, André Olivier, Jeffery Hall, Dominique Hoffmann, Dang Khoa Nguyen, Philippe Kahane, François Dubeau, Jean Gotman, Christine Rogers (20). MNI Open iEEG Atlas: normal intracranial EEG in wakefulness and sleep (processed 200 Hz atlas clips). 10.82901/nemar.nm000362
Modality: ieeg Subjects: 106 Recordings: 535 License: CC-BY-NC-4.0 Source: nemar
Metadata: Complete (100%)
106-participant iEEG dataset — MNI Open iEEG Atlas: normal intracranial EEG in wakefulness and sleep (processed 200 Hz atlas clips).
Quickstart#
Install
pip install eegdash
Access the data
from eegdash.dataset import NM000362
dataset = NM000362(cache_dir="./data")
# Get the raw object of the first recording
raw = dataset.datasets[0].raw
print(raw.info)
Filter by subject
dataset = NM000362(cache_dir="./data", subject="01")
Advanced query
dataset = NM000362(
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{nm000362,
title = {MNI Open iEEG Atlas: normal intracranial EEG in wakefulness and sleep (processed 200 Hz atlas clips)},
author = {Birgit Frauscher and Nicolás von Ellenrieder and Rina Zelmann and Irena Doležalová and Lorella Minotti and André Olivier and Jeffery Hall and Dominique Hoffmann and Dang Khoa Nguyen and Philippe Kahane and François Dubeau and Jean Gotman and Christine Rogers},
doi = {10.82901/nemar.nm000362},
url = {https://doi.org/10.82901/nemar.nm000362},
}
About This Dataset#
This is the processed part of the MNI Open iEEG Atlas: one minute of artifact-free intracranial EEG per
vigilance state (wakefulness, N2, N3, REM), plus up to 10 minutes of N2 and N3 sleep, from 106 patients.
The authors already filtered, resampled, cleaned and concatenated these data, so this is deposited as a
derivative dataset (DatasetType: derivative). The high-sampling-rate sEEG used for the normal-HFO study is
deposited separately as a raw dataset (title “MNI Open iEEG Atlas: high-sampling-rate NREM sleep sEEG”).
MNI Open iEEG Atlas: normal intracranial EEG in wakefulness and sleep (processed 200 Hz clips)
Overview
The atlas papers: wakefulness, Frauscher et al. 2018 (Brain 141:1130): “A total of 1785 channels with normal brain activity from 106 patients were identified” (abstract; the release has 1772 wake channels, see Contents); sleep, von Ellenrieder et al. 2020 (Ann Neurol 87:289): “Intracerebral electroencephalographic recordings with channels
View full README
MNI Open iEEG Atlas: normal intracranial EEG in wakefulness and sleep (processed 200 Hz clips)
Overview
The atlas papers: wakefulness, Frauscher et al. 2018 (Brain 141:1130): “A total of 1785 channels with normal brain activity from 106 patients were identified” (abstract; the release has 1772 wake channels, see Contents); sleep, von Ellenrieder et al. 2020 (Ann Neurol 87:289): “Intracerebral electroencephalographic recordings with channels displaying physiological activity from nonlesional tissue were selected from 91 patients of 3 tertiary epilepsy centers. Sections during non-rapid eye movement sleep (stages N2 and N3) and rapid eye movement sleep (stage R) were selected from the first sleep cycle” and “Results of 1,468 channels were grouped into 38 regions” (abstract).
License
CC-BY-NC-4.0
Cohort
Patients with drug-resistant focal epilepsy investigated with intracranial EEG at three tertiary epilepsy
centres (Frauscher et al. 2018). Only channels judged by the authors to record normal activity (outside the
epileptic zone and lesions, in grey matter) are included. participants.tsv holds only what the source
publishes: the source patient number (1-110; participant label sub-<3-digit number>), sex and age at time
of study (PatientInformation.csv), and the recording-centre code (first character of the source channel
names: G, M or N; the source does not name the centres). No other clinical information exists in the source
and none was added.
- Patients: 106 (58 male, 48 female; age at time of study 13-62 years; centre codes G 49, M 39, N 18;
participants.tsv). The readme: “The numbers range from 1 to 110, but there are only 106 different patients (no suitable channels were found in patients 51, 86, 95, and 105, so these patients do not appear in the database).”
Centres (Frauscher et al. 2018, Brain, Methods, “Selection of intracranial EEG recordings”): Montreal Neurological Institute and Hospital (MNI), Centre Hospitalier de l’Université de Montréal (CHUM) and Grenoble-Alpes University Hospital (CHUGA). Patients were screened “starting with the most recent patients at time of data collection (September 2015 for MNI and CHUM, April 2016 for CHUGA), and moving consecutively backward to January 2010 or earlier”. The source does not state which centre code (G, M, N) corresponds to which centre, so no mapping is given.
Implants: stereo-EEG depth electrodes (DIXI, MNI homemade, AdTech) and AdTech subdural strips/grids (source
ChannelTypeD, M, A, G). The paper (Brain, Results, Fig. 2): “stereo-EEG (commercial DIXI electrodes, home-made MNI electrodes, or Ad-Tech electrodes) or cortical grids/strips (Ad-Tech electrodes)”.Original sampling rates before the source resampled to 200 Hz (Brain, Methods): “200, 256, 512, 1000, 1024, and 2000 samples per second”; a minimum of 200 Hz was an inclusion criterion.
Selection criteria (Brain, Methods): “A channel with normal activity is defined as a channel localized in normal tissue as assessed by MRI, is located outside the seizure onset zone, does not show at any time of the circadian cycle interictal epileptic discharges (according to the clinical report of the complete implantation and to a careful investigation of one night of sleep by a board-certified electrophysiologist), and shows the absence of overt slow-wave anomaly”; contacts in white matter were excluded; recordings had to be obtained “after a minimum of 72 h after insertion of stereo-EEG electrodes or 1 week after placement of subdural grids or strips […] and at least 12 h after a generalized tonic-clonic seizure, 6 h in case of focal clinical seizures, or 2 h in case of purely electrographic seizures, and not after electrical stimulation”. Patients with large cortical malformations were excluded.
Task and states
No task or stimulus. task-wake: quiet wakefulness with eyes closed; the Brain paper (Methods) selected 60 s
“(either continuous or consecutive discontinuous >5 s segments after artefact exclusion)” of “resting wakefulness EEG
with eyes closed recorded during standardized conditions”, part of the controlled “eyes closed and eyes opened”
recordings of the clinical evaluation. task-sleepN2, task-sleepN3, task-sleepREM: sleep stages N2, N3 and R
from the first sleep cycle (von Ellenrieder et al. 2020). Reference: every channel is a bipolar derivation between
adjacent contacts of one electrode. Power-line frequency per the readme: 50 Hz for channels whose name begins with
‘G’, 60 Hz for ‘M’ or ‘N’ (PowerLineFrequency in *_ieeg.json). Amplifier/acquisition system and hardware
filters are not stated in the readmes or in the Brain paper and are left n/a.
Contents
| acq | task | Source file | Channels / subjects | Samples |
|---|---|---|---|---|
| 1min | wake (quiet wakefulness, eyes closed) | MatlabFile.mat `Data_W` | 1772 / 106 | 13600 (68 s) |
| 1min | sleepN2 | MatlabFile.mat `Data_N2` | 1468 / 91 | 13600 |
| 1min | sleepN3 | MatlabFile.mat `Data_N3` | 1468 / 91 | 13600 |
| 1min | sleepREM | MatlabFile.mat `Data_R` | 1012 / 65 | 13600 |
| 10min | sleepN2 | NREM-sleep-20min.mat `Data_N2` | 1468 / 91 | 123600 (10 min 18 s) |
| 10min | sleepN3 | NREM-sleep-20min.mat `Data_N3` | 1468 / 91 | 125200 (10 min 26 s) |
All at 200 Hz, microvolts (float32). Sleep was taken from the first sleep cycle (von Ellenrieder et al. 2020).
The 10-min N2/N3 variables are the corrected September 2020 version (changelog_September2020.txt). The
source’s EDF copies of the 1-min set (*_AllRegions.zip, one file per brain region) are kept in sourcedata/;
they match MatlabFile.mat to within 0.5 of the EDF quantisation step.
Files per recording:
- *_ieeg.vhdr/.vmrk/.eeg: BrainVision, IEEE_FLOAT_32, µV, bit-identical to the source arrays.
- *_ieeg.json: sampling rate, power-line frequency, the source processing (SoftwareFilters), reference,
electrode manufacturer, source file and variable.
*_channels.tsv: one row per bipolar channel (source names); type SEEG (depth) or ECOG (strips/grids).*_space-MNI152NLin2009aSym_electrodes.tsv+*_coordsystem.json(peracq): one row per bipolar channel at the midpoint of its two contacts, with hemisphere, region number/name, lobe and electrode type.*_events.tsv/.json: artifact-free segments (withsleep_stageW/N2/N3/R), zero buffers and end padding.sub-*_scans.tsv: list of recordings (no acquisition times exist).sourcedata/document_repository/: original release files (readmes, changelogs, Information.zip, MatlabFile.zip, NREM_sleep_20min.zip, *_AllRegions.zip, BandPowerDistribution.pdf), byte-identical with SHA-256 checksums.sourcedata/eegbrowser/: files served by the atlas’s online EEG browser (see Coordinates).
Processing done by the source authors (verbatim summary of the readmes)
1-min set (readme_MNI_Open_iEEG_Atlas.txt):
1. “All signals were resampled to 200 samples per second (unless that was the original sampling rate), after
applying a low-pass antialiasing filter at 80 Hz.”
Power-line interference reduced with an adaptive filter (harmonics estimated and subtracted; 50 Hz for channels whose name begins with ‘G’, 60 Hz for ‘M’ or ‘N’).
“Artifacts were visually detected by an experienced neurophysiologist, and excluded from the recording. This resulted in some patients having several non-consecutive segments to complete one minute of data.”
“The mean value of each segment and channel was subtracted from the corresponding segment and channel.”
“The segments were concatenated leaving a buffer time of 2 seconds of zero amplitude between segments. … All the channels were then zero padded at the end to a length of 68 seconds (13600 samples) if necessary”
10-min N2/N3 set (readme_NREMsleep.txt): the same steps without the power-line step; concatenated length
“10 min 18 seconds for stage N2 and 10 min 26 seconds for stage N3”.
Known caveats
These files are not continuous recordings. Each *_events.tsv lists every artifact-free segment
(trial_type = artifact_free_segment, with sleep_stage and segment_index), every 2-s zero buffer
(zero_buffer_between_segments) and the end padding (zero_padding_end), found as samples where all channels
of the subject are exactly 0 µV. Consecutive segments were not contiguous in the original recording. Every
buffer found is exactly 400 samples (2 s); the largest segment counts (5 in wake, 9 in 10-min N2, 13 in 10-min
N3) match the numbers stated in the readmes. Onsets are relative to the file start; there are no acquisition
dates or times. Sidecars set RecordingType: discontinuous and list the source processing in SoftwareFilters.
Coordinates: MNI space, “ICBM 2009a symmetric template (1x1x1 mm)” per the source, via nonlinear coregistration
(space-MNI152NLin2009aSym). Regions: 38 grey-matter regions (RegionInformation.csv, from a segmentation
derived from the MICCAI 2012 multi-atlas labelling template). Source inconsistency kept as published: channels
MM076LOF1 and MM076LOF2 have region “NA” in ChannelInformation.csv and 16 in MatlabFile.mat; electrodes.tsv
uses MatlabFile.mat. sourcedata/eegbrowser/ holds the template (model_mni.nii.gz), the region-label volume
(labels_mni.nii.gz) and the channel metadata (GetMeta.json) served by the atlas’s online EEG browser.
- The Brain paper abstract counts 1785 wake channels; the release has 1772. The source does not explain the
difference.
The Brain paper (Methods) states the co-registration target as the “ICBM152 2009c non-linear symmetric brain model”, while the release readme states “ICBM 2009a symmetric template” for the channel positions. The space label follows the readme of the released coordinates.
Missing stages: channels without data for a stage (NaN in the source) are absent from that recording, so the channel sets of
wake,sleepN2/N3andsleepREMof one subject can differ.
Channels and coordinates
Every channel is a bipolar derivation between adjacent contacts of one electrode. electrodes.tsv has one
row per bipolar channel, positioned at the midpoint of its two contacts, coordinates copied unchanged from
the source. Channel names are the source names. Channel type SEEG for depth electrodes (source types D = DIXI,
M = MNI homemade, A = AdTech), ECOG for subdural strips/grids (type G = AdTech).
How to load
from mne_bids import BIDSPath, read_raw_bids
bp = BIDSPath(root=".", subject="001", task="wake", acquisition="1min",
datatype="ieeg", suffix="ieeg", extension=".vhdr")
raw = read_raw_bids(bp) # bipolar channels, 200 Hz (MNE stores volts)
events = raw.annotations # artifact-free segments, 2-s zero buffers, end padding
Drop the zero_buffer_between_segments and zero_padding_end intervals before computing spectra or other
statistics.
How to cite
Cite all papers that describe the data you use: - Frauscher B, von Ellenrieder N, Zelmann R, Doležalová I, Minotti L, Olivier A, Hall J, Hoffmann D, Nguyen DK, Kahane P, Dubeau F, Gotman J. Atlas of the normal intracranial electroencephalogram: neurophysiological awake activity in different cortical areas. Brain 2018;141(4):1130-1144. doi:10.1093/brain/awy035 - Frauscher B, von Ellenrieder N, Zelmann R, Rogers C, Nguyen DK, Kahane P, Dubeau F, Gotman J. High-Frequency Oscillations in the Normal Human Brain. Ann Neurol 2018;84(3):374-385. doi:10.1002/ana.25304 - von Ellenrieder N, Gotman J, Zelmann R, Rogers C, Nguyen DK, Kahane P, Dubeau F, Frauscher B. How the Human Brain Sleeps: Direct Cortical Recordings of Normal Brain Activity. Ann Neurol 2020;87(2):289-301. doi:10.1002/ana.25651
Source: MNI Open iEEG Atlas, Montreal Neurological Institute, https://mni-open-ieegatlas.research.mcgill.ca/
Ethics
Frauscher et al. 2018 (Brain 141:1130), Methods: “Ethical approval was granted at the MNI as lead ethics organization (REB vote: MUHC-15-950).” The data were recorded during clinical presurgical evaluation of drug-resistant focal epilepsy at three tertiary epilepsy centres.
Funding
Frauscher et al. 2018 (Brain 141:1130), Funding: “This work was supported by the Savoy Epilepsy Foundation (project grant to B.F. and post-doctoral fellowship to R.Z.), the Botterell Powell’s Foundation (grant to B.F.), and the Canadian Institute of Health Research (grant FDN-143208 to J.G.).” Crossref funder record of von Ellenrieder et al. 2020 (doi:10.1002/ana.25651): Canadian Institutes of Health Research “FDN-143208 (JG)”; Fonds de Recherche du Québec - Santé “Chercheur-boursier clinicien Junior 2 (BF)”.
Participants
See Cohort. participants.tsv also flags whether a patient is in the 1-min atlas and in the 10-min NREM set, and
lists the hemispheres, electrode types and channel count of the patient’s atlas channels.
Privacy
The source files were already anonymised by the authors (EDF headers carry “X” placeholders and the date
01-JAN-1970; no names, birth dates or recording dates). A byte-level review of every converted file and of
sourcedata/ found no participant identifiers. The only personal names are author attributions (the
HFODetector.m author line and the PDF author field). The NIfTI volumes are brain templates, not participant
images.
Conversion
Conversion (iEEG-NEMAR campaign, lane F, 2026-10-06) is a lossless repack: every sample in the BrainVision
.eeg files is bit-identical to the source arrays (606/606 recordings verified, raw bytes compared). No
filtering, resampling, re-referencing, rescaling or channel selection was done. Channels absent in a stage
(NaN columns in the source) are omitted from that recording, never filled. All original files are kept
byte-identical under sourcedata/ with SHA-256 checksums.
NEMAR Metadata#
[](https://doi.org/10.82901/nemar.nm000362) # MNI Open iEEG Atlas: normal intracranial EEG in wakefulness and sleep (processed 200 Hz clips) ## Overview This is the processed part of the MNI Open iEEG Atlas: one minute of artifact-free intracranial EEG per vigilance state (wakefulness, N2, N3, REM), plus up to 10 minutes of N2 and N3 sleep, from 106 patients. The authors already filtered, resampled, cleaned and concatenated these data, so this is deposited as a derivative dataset (DatasetType: derivative). The high-sampling-rate sEEG used for the normal-HFO study is deposited separately as a raw dataset (title “MNI Open iEEG Atlas: high-sampling-rate NREM sleep sEEG”). The atlas papers: wakefulness, Frauscher et al. 2018 (Brain 141:1130): “A total of 1785 channels with normal brain activity from 106 patients were identified” (abstract; the release has 1772 wake channels, see Contents); sleep, von Ellenrieder et al. 2020 (Ann Neurol 87:289): “Intracerebral electroencephalographic recordings with channels displaying physiological activity from nonlesional tissue were selected from 91 patients of 3 tertiary epilepsy centers. Sections during non-rapid eye movement sleep (stages N2 and N3) and rapid eye movement sleep (stage R) were selected from the first sleep cycle” and “Results of 1,468 channels were grouped into 38 regions” (abstract). ## License CC-BY-NC-4.0 ## Cohort Patients with drug-resistant focal epilepsy investigated with intracranial EEG at three tertiary epilepsy centres (Frauscher et al. 2018). Only channels judged by the authors to record normal activity (outside the epileptic zone and lesions, in grey matter) are included. participants.tsv holds only what the source publishes: the source patient number (1-110; participant label sub-<3-digit number>), sex and age at time of study (PatientInformation.csv), and the recording-centre code (first character of the source channel names: G, M or N; the source does not name the centres). No other clinical information exists in the source and none was added. - Patients: 106 (58 male, 48 female; age at time of study 13-62 years; centre codes G 49, M 39, N 18;
participants.tsv). The readme: “The numbers range from 1 to 110, but there are only 106 different patients (no suitable channels were found in patients 51, 86, 95, and 105, so these patients do not appear in the database).”
Centres (Frauscher et al. 2018, Brain, Methods, “Selection of intracranial EEG recordings”): Montreal Neurological Institute and Hospital (MNI), Centre Hospitalier de l’Université de Montréal (CHUM) and Grenoble-Alpes University Hospital (CHUGA). Patients were screened “starting with the most recent patients at time of data collection (September 2015 for MNI and CHUM, April 2016 for CHUGA), and moving consecutively backward to January 2010 or earlier”. The source does not state which centre code (G, M, N) corresponds to which centre, so no mapping is given.
Implants: stereo-EEG depth electrodes (DIXI, MNI homemade, AdTech) and AdTech subdural strips/grids (source ChannelType D, M, A, G). The paper (Brain, Results, Fig. 2): “stereo-EEG (commercial DIXI electrodes, home-made MNI electrodes, or Ad-Tech electrodes) or cortical grids/strips (Ad-Tech electrodes)”.
Original sampling rates before the source resampled to 200 Hz (Brain, Methods): “200, 256, 512, 1000, 1024, and 2000 samples per second”; a minimum of 200 Hz was an inclusion criterion.
Selection criteria (Brain, Methods): “A channel with normal activity is defined as a channel localized in normal tissue as assessed by MRI, is located outside the seizure onset zone, does not show at any time of the circadian cycle interictal epileptic discharges (according to the clinical report of the complete implantation and to a careful investigation of one night of sleep by a board-certified electrophysiologist), and shows the absence of overt slow-wave anomaly”; contacts in white matter were excluded; recordings had to be obtained “after a minimum of 72 h after insertion of stereo-EEG electrodes or 1 week after placement of subdural grids or strips […] and at least 12 h after a generalized tonic-clonic seizure, 6 h in case of focal clinical seizures, or 2 h in case of purely electrographic seizures, and not after electrical stimulation”. Patients with large cortical malformations were excluded.
## Task and states No task or stimulus. task-wake: quiet wakefulness with eyes closed; the Brain paper (Methods) selected 60 s “(either continuous or consecutive discontinuous >5 s segments after artefact exclusion)” of “resting wakefulness EEG with eyes closed recorded during standardized conditions”, part of the controlled “eyes closed and eyes opened” recordings of the clinical evaluation. task-sleepN2, task-sleepN3, task-sleepREM: sleep stages N2, N3 and R from the first sleep cycle (von Ellenrieder et al. 2020). Reference: every channel is a bipolar derivation between adjacent contacts of one electrode. Power-line frequency per the readme: 50 Hz for channels whose name begins with ‘G’, 60 Hz for ‘M’ or ‘N’ (PowerLineFrequency in *_ieeg.json). Amplifier/acquisition system and hardware filters are not stated in the readmes or in the Brain paper and are left n/a. ## Contents | acq | task | Source file | Channels / subjects | Samples | |---|—|---|—|---| | 1min | wake (quiet wakefulness, eyes closed) | MatlabFile.mat Data_W | 1772 / 106 | 13600 (68 s) | | 1min | sleepN2 | MatlabFile.mat Data_N2 | 1468 / 91 | 13600 | | 1min | sleepN3 | MatlabFile.mat Data_N3 | 1468 / 91 | 13600 | | 1min | sleepREM | MatlabFile.mat Data_R | 1012 / 65 | 13600 | | 10min | sleepN2 | NREM-sleep-20min.mat Data_N2 | 1468 / 91 | 123600 (10 min 18 s) | | 10min | sleepN3 | NREM-sleep-20min.mat Data_N3 | 1468 / 91 | 125200 (10 min 26 s) | All at 200 Hz, microvolts (float32). Sleep was taken from the first sleep cycle (von Ellenrieder et al. 2020). The 10-min N2/N3 variables are the corrected September 2020 version (changelog_September2020.txt). The source’s EDF copies of the 1-min set (*_AllRegions.zip, one file per brain region) are kept in sourcedata/; they match MatlabFile.mat to within 0.5 of the EDF quantisation step. Files per recording: - *_ieeg.vhdr/.vmrk/.eeg: BrainVision, IEEE_FLOAT_32, µV, bit-identical to the source arrays. - *_ieeg.json: sampling rate, power-line frequency, the source processing (SoftwareFilters), reference,
electrode manufacturer, source file and variable.
*_channels.tsv: one row per bipolar channel (source names); type SEEG (depth) or ECOG (strips/grids).
*_space-MNI152NLin2009aSym_electrodes.tsv + *_coordsystem.json (per acq): one row per bipolar channel at the midpoint of its two contacts, with hemisphere, region number/name, lobe and electrode type.
*_events.tsv/.json: artifact-free segments (with sleep_stage W/N2/N3/R), zero buffers and end padding.
sub-*_scans.tsv: list of recordings (no acquisition times exist).
sourcedata/document_repository/: original release files (readmes, changelogs, Information.zip, MatlabFile.zip, NREM_sleep_20min.zip, *_AllRegions.zip, BandPowerDistribution.pdf), byte-identical with SHA-256 checksums.
sourcedata/eegbrowser/: files served by the atlas’s online EEG browser (see Coordinates).
## Processing done by the source authors (verbatim summary of the readmes) 1-min set (readme_MNI_Open_iEEG_Atlas.txt): 1. “All signals were resampled to 200 samples per second (unless that was the original sampling rate), after
applying a low-pass antialiasing filter at 80 Hz.”
Power-line interference reduced with an adaptive filter (harmonics estimated and subtracted; 50 Hz for channels whose name begins with ‘G’, 60 Hz for ‘M’ or ‘N’).
“Artifacts were visually detected by an experienced neurophysiologist, and excluded from the recording. This resulted in some patients having several non-consecutive segments to complete one minute of data.”
“The mean value of each segment and channel was subtracted from the corresponding segment and channel.”
“The segments were concatenated leaving a buffer time of 2 seconds of zero amplitude between segments. … All the channels were then zero padded at the end to a length of 68 seconds (13600 samples) if necessary”
10-min N2/N3 set (readme_NREMsleep.txt): the same steps without the power-line step; concatenated length “10 min 18 seconds for stage N2 and 10 min 26 seconds for stage N3”. ## Known caveats These files are not continuous recordings. Each *_events.tsv lists every artifact-free segment (trial_type = artifact_free_segment, with sleep_stage and segment_index), every 2-s zero buffer (zero_buffer_between_segments) and the end padding (zero_padding_end), found as samples where all channels of the subject are exactly 0 µV. Consecutive segments were not contiguous in the original recording. Every buffer found is exactly 400 samples (2 s); the largest segment counts (5 in wake, 9 in 10-min N2, 13 in 10-min N3) match the numbers stated in the readmes. Onsets are relative to the file start; there are no acquisition dates or times. Sidecars set RecordingType: discontinuous and list the source processing in SoftwareFilters. Coordinates: MNI space, “ICBM 2009a symmetric template (1x1x1 mm)” per the source, via nonlinear coregistration (space-MNI152NLin2009aSym). Regions: 38 grey-matter regions (RegionInformation.csv, from a segmentation derived from the MICCAI 2012 multi-atlas labelling template). Source inconsistency kept as published: channels MM076LOF1 and MM076LOF2 have region “NA” in ChannelInformation.csv and 16 in MatlabFile.mat; electrodes.tsv uses MatlabFile.mat. sourcedata/eegbrowser/ holds the template (model_mni.nii.gz), the region-label volume (labels_mni.nii.gz) and the channel metadata (GetMeta.json) served by the atlas’s online EEG browser. - The Brain paper abstract counts 1785 wake channels; the release has 1772. The source does not explain the
difference.
The Brain paper (Methods) states the co-registration target as the “ICBM152 2009c non-linear symmetric brain model”, while the release readme states “ICBM 2009a symmetric template” for the channel positions. The space label follows the readme of the released coordinates.
Missing stages: channels without data for a stage (NaN in the source) are absent from that recording, so the channel sets of wake, sleepN2/N3 and sleepREM of one subject can differ.
## Channels and coordinates Every channel is a bipolar derivation between adjacent contacts of one electrode. electrodes.tsv has one row per bipolar channel, positioned at the midpoint of its two contacts, coordinates copied unchanged from the source. Channel names are the source names. Channel type SEEG for depth electrodes (source types D = DIXI, M = MNI homemade, A = AdTech), ECOG for subdural strips/grids (type G = AdTech). ## How to load ```python from mne_bids import BIDSPath, read_raw_bids bp = BIDSPath(root=”.”, subject=”001”, task=”wake”, acquisition=”1min”,
datatype=”ieeg”, suffix=”ieeg”, extension=”.vhdr”)
raw = read_raw_bids(bp) # bipolar channels, 200 Hz (MNE stores volts) events = raw.annotations # artifact-free segments, 2-s zero buffers, end padding ``` Drop the zero_buffer_between_segments and zero_padding_end intervals before computing spectra or other statistics. ## How to cite Cite all papers that describe the data you use: - Frauscher B, von Ellenrieder N, Zelmann R, Doležalová I, Minotti L, Olivier A, Hall J, Hoffmann D, Nguyen DK, Kahane P, Dubeau F, Gotman J. Atlas of the normal intracranial electroencephalogram: neurophysiological awake activity in different cortical areas. Brain 2018;141(4):1130-1144. doi:10.1093/brain/awy035 - Frauscher B, von Ellenrieder N, Zelmann R, Rogers C, Nguyen DK, Kahane P, Dubeau F, Gotman J. High-Frequency Oscillations in the Normal Human Brain. Ann Neurol 2018;84(3):374-385. doi:10.1002/ana.25304 - von Ellenrieder N, Gotman J, Zelmann R, Rogers C, Nguyen DK, Kahane P, Dubeau F, Frauscher B. How the Human Brain Sleeps: Direct Cortical Recordings of Normal Brain Activity. Ann Neurol 2020;87(2):289-301. doi:10.1002/ana.25651 Source: MNI Open iEEG Atlas, Montreal Neurological Institute, https://mni-open-ieegatlas.research.mcgill.ca/ ## Ethics Frauscher et al. 2018 (Brain 141:1130), Methods: “Ethical approval was granted at the MNI as lead ethics organization (REB vote: MUHC-15-950).” The data were recorded during clinical presurgical evaluation of drug-resistant focal epilepsy at three tertiary epilepsy centres. ## Funding Frauscher et al. 2018 (Brain 141:1130), Funding: “This work was supported by the Savoy Epilepsy Foundation (project grant to B.F. and post-doctoral fellowship to R.Z.), the Botterell Powell’s Foundation (grant to B.F.), and the Canadian Institute of Health Research (grant FDN-143208 to J.G.).” Crossref funder record of von Ellenrieder et al. 2020 (doi:10.1002/ana.25651): Canadian Institutes of Health Research “FDN-143208 (JG)”; Fonds de Recherche du Québec - Santé “Chercheur-boursier clinicien Junior 2 (BF)”. ## Participants See Cohort. participants.tsv also flags whether a patient is in the 1-min atlas and in the 10-min NREM set, and lists the hemispheres, electrode types and channel count of the patient’s atlas channels. ## Privacy The source files were already anonymised by the authors (EDF headers carry “X” placeholders and the date 01-JAN-1970; no names, birth dates or recording dates). A byte-level review of every converted file and of sourcedata/ found no participant identifiers. The only personal names are author attributions (the HFODetector.m author line and the PDF author field). The NIfTI volumes are brain templates, not participant images. ## Conversion Conversion (iEEG-NEMAR campaign, lane F, 2026-10-06) is a lossless repack: every sample in the BrainVision .eeg files is bit-identical to the source arrays (606/606 recordings verified, raw bytes compared). No filtering, resampling, re-referencing, rescaling or channel selection was done. Channels absent in a stage (NaN columns in the source) are omitted from that recording, never filled. All original files are kept byte-identical under sourcedata/ with SHA-256 checksums.
License: CC-BY-NC-4.0
Authors:
Birgit Frauscher
Nicolás von Ellenrieder
Rina Zelmann
Irena Doležalová
Lorella Minotti
… and 8 more
Versions:
Version |
DOI |
Released |
|---|---|---|
|
Cohort#
Dataset Statistics#
Age distribution by gender (n=106, range 13–62 yr, mean 33.4 yr)
Sex composition
Channel counts (ch)
Sampling frequencies: 200.0 Hz (n=535 recordings)
Total recording duration: 38 h
Signal · Electrodes & live trace#
Live trace viewer — sub-031 · task-sleepN2
Showing one representative recording out of
106 subjects and 535 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 · 1 sensors — 1 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
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.
Full dataset metadata table
Dataset ID |
|
Title |
MNI Open iEEG Atlas: normal intracranial EEG in wakefulness and sleep (processed 200 Hz atlas clips) |
Author (year) |
— |
Canonical |
— |
Importable as |
|
Year |
20 |
Authors |
Birgit Frauscher, Nicolás von Ellenrieder, Rina Zelmann, Irena Doležalová, Lorella Minotti, André Olivier, Jeffery Hall, Dominique Hoffmann, Dang Khoa Nguyen, Philippe Kahane, François Dubeau, Jean Gotman, Christine Rogers |
License |
CC-BY-NC-4.0 |
Citation / DOI |
|
Source links |
OpenNeuro | NeMAR | Source URL |
Copy-paste BibTeX
@dataset{nm000362,
title = {MNI Open iEEG Atlas: normal intracranial EEG in wakefulness and sleep (processed 200 Hz atlas clips)},
author = {Birgit Frauscher and Nicolás von Ellenrieder and Rina Zelmann and Irena Doležalová and Lorella Minotti and André Olivier and Jeffery Hall and Dominique Hoffmann and Dang Khoa Nguyen and Philippe Kahane and François Dubeau and Jean Gotman and Christine Rogers},
doi = {10.82901/nemar.nm000362},
url = {https://doi.org/10.82901/nemar.nm000362},
}
API Reference#
eegdash.datasetEEGDashDataset- class eegdash.dataset.NM000362(cache_dir: str, query: dict | None = None, s3_bucket: str | None = None, **kwargs)[source]#
MNI Open iEEG Atlas: normal intracranial EEG in wakefulness and sleep (processed 200 Hz atlas clips)
- Study:
nm000362(NeMAR)- Author (year):
—
- Canonical:
—
Also importable as:
NM000362.Modality:
ieeg; Subject type:Unknown. Subjects: 106; recordings: 535; tasks: 4.- 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
Notes
Each item is a recording; recording-level metadata are available via
dataset.description.querysupports MongoDB-style filters on fields inALLOWED_QUERY_FIELDSand is combined with the dataset filter. Dataset-specific caveats are not provided in the summary metadata.References
OpenNeuro dataset: https://openneuro.org/datasets/nm000362 NeMAR dataset: https://nemar.org/dataexplorer/detail?dataset_id=nm000362 DOI: https://doi.org/10.82901/nemar.nm000362
Examples
>>> from eegdash.dataset import NM000362 >>> dataset = NM000362(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.
BaseDataset from braindecode — windowed via create_windows_from_events.braindecodeDataLoader; supports parallel workers and on-the-fly augmentations.pytorch
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 nm000362 to reproduce the tutorial on this dataset.
Citation
Birgit Frauscher, Nicolás von Ellenrieder, Rina Zelmann, Irena Doležalová, Lorella Minotti, … (20). MNI Open iEEG Atlas: normal intracranial EEG in wakefulness and sleep (processed 200 Hz atlas clips). 10.82901/nemar.nm000362
Provenance
¹Contributed to nemar in BIDS format.
²Curated & ingested by the EEGDash catalog; see CITATION.cff for canonical reference.
³Persistent identifier: 10.82901/nemar.nm000362.
See Also#
eegdash.dataset.EEGDashDataseteegdash.dataset