Adjacent bipolar subtraction raises the mean 30–100 Hz SOZ slope contrast versus common average in matched pairs from 30-second sections in 35 HUP SEEG patients
We tested whether the reference method changes a SEEG spectral slope contrast. We used 30-second sections from two public interictal runs in each of 36 HUP patients. The primary comparison included 35 patients with enough pure seizure-onset and other contact pairs. The feature was the negative log-log power slope from 30 to 100 Hz, with 55 to 65 Hz excluded. Pair membership was fixed across references. The mean change from common average to adjacent bipolar subtraction was 0.178 slope units. The patient bootstrap 95 percent confidence interval was 0.070 to 0.286. The SOZ-minus-other contrast changed sign in 4 of 35 patients. The package preserves spectra, labels, code, results, source versions, hashes, and retrieval instructions. The evidence applies to these short sections and this slope feature. It does not establish a clinical test or a measure of excitation and inhibition.
stereoelectroencephalography · seizure-onset zone · spectral slope · bipolar reference · common average · reference methods
References
3- usesdoi:10.18112/openneuro.ds004100.v1.1.3
Source signals and contact labels came from this exact OpenNeuro snapshot.
- doi:10.1016/j.clinph.2006.05.032
- doi:10.1002/advs.202200887
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesMean bipolar SEEG slope rank agreement was lower between 30–100 Hz and 80–150 Hz fits than between recordings in 24-second samples from 35 HUP participants
Used its published description of the 30 to 100 Hz SEEG slope analysis to choose the lower fit range. The current analysis kept bipolar subtraction fixed and compared two fit ranges with two recordings. It used different short sections from the same HUP source cohort.
- usesThe real cross-spectrum term raises the mean 30–100 Hz SOZ slope contrast compared with summed endpoint powers in two 30-second interictal sections from 35 HUP SEEG patients
We used its 72 archived power-spectrum files, pinned source recipes, metadata, contact selection rules, and earlier reference result. We added summed endpoint power controls and direct cross-spectra for a new patient-level slope decomposition.
- usesDivision by a 30–70 Hz prediction reversed the mean seizure onset label relation for 80–150 Hz bipolar SEEG power in short samples from 34 patients
Its spectral slope result guided the question about high gamma power and a low-band prediction. Its HUP180 label audit guided the planned patient exclusion. We did not treat this same-dataset follow-up as an independent replication.
- usesContact holdout gave higher SOZ AUC than patient holdout for a fixed random forest in 36 short SEEG records
We used its 72 public SEEG spectra, clinical channel tables, source byte records, and reader. These files enabled fixed feature extraction, model split tests, and raw source replay.