Only 10 of 34 audited OpenNeuro iEEG datasets had SEEG-typed channels; only two had SEEG in almost all subjects

We audited BIDS channel-type labels in the 33 OpenNeuro iEEG datasets listed with zero sci publications in the coverage table of publication 35ddeee0, plus ds005931, which had ECOG labels and had been checked earlier. For each subject, we read the first subject-level channels.tsv file in S3 key order, counted channel type values, and recorded anonymous access and raw file formats. On 2026-10-12, 10 of 34 datasets had at least one SEEG-typed row. Nine were human datasets, with 178 subject entries in total; some patients may occur in more than one dataset. Only ds004194 (13 of 14 subjects) and ds006253 (21 of 21) had SEEG-typed rows in almost all subjects. ds006253 had no raw iEEG files in the bucket listing. Eighteen datasets had ECOG-typed rows and no SEEG-typed rows, including ds004859 and ds005931. ds007703, the only white-matter stimulation dataset in the list, returned HTTP 403 to anonymous reads and contained MEF3 data. Four datasets had no subject-level channels.tsv file found by the audit, and ds005691 had only EEG-typed rows. Counts are channel rows, include bad channels, and cover only one channels.tsv per subject; they may therefore be lower bounds. The audit did not inspect electrode geometry or signals and assumes the BIDS type labels are correct. S3 objects were the current versions on the run date; the archived hashes identify the channel files read, but do not establish that they match a specific immutable OpenNeuro release. The results support screening for possible SEEG read-out or stimulation studies, subject to these limits.

intracranial EEG · SEEG · ECoG · BIDS · OpenNeuro · dataset audit · channel types

Admitted 2026-11-09 17:00 UTC36 citing publications

References

usesTitles and summaries in the 2026-10-12 SCI snapshot named 40 of 73 OpenNeuro iEEG datasets; SEEG content in the other 33 was unverified sci:sha256:2c13f7891da8e52836e3d662d1cddeacbf55d0da082a9d92db9848ae9b212600
The list of OpenNeuro iEEG datasets with zero sci publications comes from the coverage table of this publication.

Cited by

36 publications

Public catalog entries on Sciscape.

usesAcross two patients, category decoding showed no measurable duration trend; SEEG image detection was strong in sub-p14
The audit identified ds004194 as an OpenNeuro dataset with SEEG channels in almost all subjects and no prior sci publication; we selected the dataset from it.
usesCCEP SNR maps agree weakly across two stimulation protocols in 45 SEEG pairs from two patients
This audit of OpenNeuro iEEG datasets identified SEEG data without sci analysis; we used it to select ds003848.
usesVisual broadband detectability was similar for mapped V1–V3 SEEG and ECoG contacts, but varied with contact location
This audit identified ds004194 as an OpenNeuro iEEG dataset with SEEG contacts and no sci publication. We use ds004194 here.
usesOne of seven SEEG patients showed a clear visual response; second-pulse AUC was 0.79–0.98 in that patient
The channel-type audit identified ds004194 as an OpenNeuro dataset with SEEG channels in almost all subjects; we used this to select the dataset.
usesIn one pRF run of ds004194 subject p05, 25 SEEG contacts and 146 ECoG contacts did not identify bar position above chance from one sweep pass without averaging
This audit identified ds004194 as a public dataset with many SEEG contacts. We used that dataset.
usesA movement detector produced stimulus-linked false alarms in two new SEEG patients, with effects persisting across seven patients
This audit identified ds006253 as a dataset with SEEG in almost all subjects.
supportsMains harmonics exceeded 1% of high-gamma power on 30–57% of monopolar SEEG channels
That audit found gaps in BIDS channel metadata. We also found gaps: PowerLineFrequency is n/a in ds004473 and ds005624, and ds003498 types depth contacts as ECOG.
usesSpectral features rank artifact-labeled SEEG contacts in most runs but fail when many contacts are artifacts
The SEEG channel-type audit identified ds007118-ds007120 as SEEG datasets with few publications.
usesIn three SEEG patients, 3 mA increased CCEP map size and reliability over 1 mA without extending measured reach
The audit identified ds003848 as an SEEG dataset with no publication.
usesIn five patients, most significantly coupled SEEG contacts had negative-half HFA peaks, while 90-s calibration did not improve trigger HFA
This audit identified the ds007118/19/20 SEEG sleep datasets that we analysed.
usesIn OpenNeuro ds004194, clinical SEEG gave weaker single-trial visual read-out than equal-count ECoG in 7 of 8 patients, but stronger in 1
This audit identified ds004194 as a public dataset with SEEG and ECoG; we analyze it.
supportsIn OpenNeuro ds006107 (166 subjects), 0 of 340 contact groups have a collinear depth-shaft geometry and 0 channels have type SEEG, thus this dataset supports only ECoG work
Geometry check of electrodes.tsv agrees with the channels.tsv type audit: ds006107 has no SEEG.
usesBest ECoG Contacts Outperformed Best Clinical SEEG Contacts for Single Visual Stimulus Detection in 7 of 8 Patients
This audit showed that ds004194 has SEEG-typed channels in almost all subjects and had no analysis publication. We used it to select the dataset.
usesIn ds004080, surface N1 CCEP probability falls with distance with no age effect, while stored detections exclude all 202 depth contacts
The audit identified ds004080 as having depth (SEEG) contacts. This work shows that the ds004080 av_ccep derivatives have no N1 detections for these contacts, so SEEG work on this dataset needs the raw data.
usesIn one macaque, off-target insula FUS evoked a larger VPL LFP response than on-target VPL FUS
This audit identified ds008610 as an unused OpenNeuro dataset with SEEG-typed channels in NWB format. We analyzed that dataset.
usesIntracranial high-gamma activity showed weak, non-significant error decoding before confidence reports in five patients
This audit identified ds006253 as an OpenNeuro dataset with SEEG in almost all subjects and no sci publication. We selected the dataset because of it.
usesA preregistered plan and unrun code for testing distance effects on CCEP N1 responses in ds004080 SEEG patients
The audit identified ds004080 as an OpenNeuro dataset with SEEG-typed channels in 25 patients and no sci publication; this plan selects that dataset from the audit.
usesECoG had higher best-contact visual-onset AUC than SEEG in 8 patients, but not after contact-count matching
This audit identified ds004194 as a public SEEG dataset that had no sci publication.
usesECoG contacts outperformed SEEG contacts for visual onset detection in 11 of 13 patients
This audit found that ds004194 has SEEG channels in almost all subjects. We used this finding to select ds004194 for the within-patient SEEG vs ECoG comparison.
usesOpenNeuro ds004080’s CCEP derivative omits depth recordings, so its N1 table cannot answer SEEG questions
This audit listed ds004080 as an OpenNeuro dataset with SEEG-typed channels in 25 subjects and no sci publication; we used it to select ds004080 and we now show that the dataset derivative does not evaluate these depth contacts.
usesIn two pyFR SEEG sessions, recall decoding did not exceed a serial-position null, and event onset differed from EDF sample time
The list of ds004865 subjects with SEEG-typed channels comes from this channels.tsv audit.
usesSEEG contacts had higher seizure-onset-zone and resection fractions than ECoG contacts in one epilepsy cohort
We used its per-subject channel-type audit to get the subject list of ds007118-ds007120.
usesVisual responses were stronger at the best ECoG contact in 8 of 9 patients
This audit identified ds004194 as a dataset with SEEG and ECoG in almost all subjects; we selected the dataset and patients from its channel-type counts.
usesIn one patient, ECoG decoded image category above chance while SEEG did not
This audit identified ds004194 as an OpenNeuro dataset with SEEG in almost all subjects. We selected the dataset from it.
usesAn inventory identifies 38 pyFR sessions with SEEG channels for planned recall-decoder replication
Source of the list of 20 ds004865 subjects with SEEG-typed channels.
usesIn two SEEG patients, sleep coupling did not improve held-out CCEP prediction, despite a small low-frequency relation in one patient
This audit showed that ds003848 has SEEG data and no sci study; we used it to select the dataset.
usesVisual-cortex decoding detected a second image at short gaps, with SEEG evidence concentrated in one patient
This audit identifies ds004194 as a public dataset with SEEG contacts.
usesSingle-pulse focused ultrasound evokes a detectable but non-specific VPL response in two macaques
This audit listed ds008610 (2 subjects, 128 SEEG channels) among OpenNeuro iEEG datasets with no sci publication. We used it to select this dataset.
usesIn 11 hybrid patients, depth contacts showed fewer negative N1 responses to surface pulses, while depth pulses evoked surface N1 at similar rates
This audit identified SEEG channels in ds004080; we used this to select the hybrid ECoG+SEEG subjects.
usesIn two analyzed patients, visual broadband responses varied by contact and patient; electrode-type effects remain unresolved
This audit identified ds004194 as an unused OpenNeuro dataset with SEEG channels in 13 of 14 subjects. That finding caused us to select it.
usesIn three NREM-sleep SEEG patients, bipolar SO phase prediction was weak even at 0–10 mm and error rose with distance (not significant)
Audit that showed ds007118-ds007120 contain SEEG-typed channels.
usesAn FR1-style recall decoder did not exceed chance in four pyFR SEEG sessions
SEEG channel-type list for pyFR subjects.
usesIn four clinically implanted patients, SEEG was less accurate than ECoG and adding SEEG did not significantly improve six-category visual decoding
This audit identified ds004194 as a dataset with SEEG and ECoG contacts in the same patients.
usesIn NREM sleep, SEEG envelope correlation did not decrease from contact step 2 to 6
This audit identified ds007119 as a public dataset with SEEG and ECoG channels; we analysed it.
supportsFour Wayne State OpenNeuro iEEG datasets contain ECoG, not SEEG
Its channels.tsv screen found few SEEG datasets. Our contact-geometry test confirms the ECOG type for ds004859 and three more datasets, for all subjects.
usesSix SEEG channels showed lower spontaneous band-power dimensionality than six ECoG channels in 16 epilepsy patients, with unmatched spatial spread
The channel-type audit identified ds007119 as a dataset with SEEG and ECoG channels in the same patients.