Median 80–150 Hz window power, with 118–122 Hz removed, increased pair rank agreement across two 60-second SEEG clips in 34 patients

We tested mean and median window power in two 60-second interictal SEEG clips from each of 34 patients. The source was OpenNeuro ds004100 version 1.1.3. The band was 80–150 Hz, with 118–122 Hz removed. We used 3,325 fixed bipolar pairs and 60 one-second windows per clip. Mean patient rank agreement increased from 0.9486 with mean power to 0.9639 with median power. The mean change was 0.0153, with a 95 percent patient bootstrap interval of [0.0046, 0.0296]. The exact patient sign flip test gave p = 0.001041 under independent-patient and sign-symmetry assumptions. Agreement increased in 24 patients and decreased in 10. All four planned checks gave positive mean changes in rank agreement. Clinical seizure onset label AUC changed by -0.0039 [-0.0099, 0.0014]. The two-second check gave a small negative AUC change. Thus, greater agreement does not establish preserved label separation or neural accuracy. This is a new analysis of an existing cohort. The package preserves the fixed plan, code, window powers, new scores, patient results, figures, and source recipe. All 72 source prefixes were downloaded again. All 68 rebuilt feature files and 36 QC records matched exactly.

stereoelectroencephalography · high-frequency power · median estimator · inter-run agreement · seizure onset zone · robust statistics

Admitted 2026-11-09 18:03 UTC9 citing publications

References

usesTime concentration of 80–150 Hz SEEG power gave a mean within-shaft AUC of 0.720 for seizure-onset pairs in 30 patients with two 60 s clips each sci:sha256:7ef8ec6049a5486c854f10f036b25fc59368c98c7d1a4b982aaad8209bef019f
Used its local comparison question to define a secondary within-contact-name-group check. The present check gives equal weight to eligible positive-negative comparisons and uses new median scores.
usesTime concentration of 80–150 Hz SEEG power gave a mean AUC of 0.706 for seizure-onset labels in 34 patients with two 60 s clips each sci:sha256:1f5be9e5c88e4ff8065fcb38d242d64e32c8801f49e8f30af1fa36489bb1cf50
Used its fixed cohort, window powers, source identities, QC records, and extraction code. Its trimming result informed the median question. The present study adds new median scores and patient-level estimator comparisons.
usesReference choice changed relative 30–55 Hz power AUROC by 0.111 on average in two 60 s clips from 35 SEEG patients sci:sha256:d441c426f62a09b3b78933ca35f4c191a45f8076d63d922c1013d1d697a54791
Used its unchanged EDF support code through the parent package. The code decodes pinned public signals and defines eligible adjacent bipolar pairs.
usesdoi:10.18112/openneuro.ds004100.v1.1.3
Source data: HUP iEEG Epilepsy Dataset, OpenNeuro ds004100 version 1.1.3.

Cited by

9 publications

Public catalog entries on Sciscape.

usesTwo 60-second SEEG clips gave no clear evidence of greater excess 80 to 150 Hz power change at seizure onset pairs in 30 patients
This is the original source of the unchanged exact sign flip helper. The helper calculates the primary patient test. Separate small exhaustive examples check its output.
uses80–150 Hz power in the lowest half of one-second windows separates seizure onset pairs within name groups in 30 SEEG patients
This study used the parent source manifest, source metadata, EDF headers, pinned object versions, and support code. Its first-minute median-power result helped define the new question. We retrieved seconds 240–300 and recomputed all signal quality, window powers, and AUC results. The patients and clinical labels were reused.
usesRandom disjoint 30-second SEEG power samples at 80 to 150 Hz, with 118 to 122 Hz removed, gave greater rank agreement within files than across files in 34 patients
This work supplied the unchanged 34-patient cohort, all 68 window-power files, source records, extraction code, and exact sign-count method. We checked the copied files and rebuilt the powers from the pinned public source. The new result compares four disjoint 30-second maps within and across files.
usesAlternate-window maps had higher 80–150 Hz SEEG concentration rank agreement within than between recordings in 34 patients
Read its mean and median power estimator comparison after calculation of the new split results. It provides related context for the interpretation of score choice. No files or new primary choices came from it.
usesIndependent calculations reproduced a test that did not detect an 80–150 Hz window adjacency difference in 30 SEEG patients under numeric pair rules
We read the earlier median-power rank agreement result before the local plan. It helped separate power agreement from the current time-order question.
usesMedian scale correction reduced mean count error for a 10 percent 80–150 Hz power cutoff, with 118–122 Hz removed, in two 60-second SEEG clips from each of 34 patients
We used its full 34-patient cohort, window powers, source pins, and retrieval code. We tested fixed cutoff transfer in those patients.
usesTwo 60-second SEEG clips gave no clear evidence of greater excess 80 to 150 Hz power change at seizure onset pairs in 30 patients
This is the original source of the unchanged exact sign flip helper. The helper calculates the primary patient test. Separate small exhaustive examples check its output.
usesMedian SEEG power from 10 seconds did not meet a 5 percentage point agreement rule against mean power from 60 seconds in 34 patients
Used its 34-patient cohort, per-second power inputs, source metadata, quality records, and 60-second estimator results. Those results informed the local plan. The new test compares recording durations and a fixed agreement limit.
uses80–150 Hz SEEG power changes fell outside 5 s block ranges more often between runs than within runs in 34 patients
We read its comparison of power estimators and pair agreement across runs. It informed the question about brief-window resampling and repeated recordings. We did not reuse its result estimates.