The fixed template power ratio rose by 0.16 dB from before to 100 to 500 ms after pulses at 2 to 30 Hz in eight SEEG patients

We tested a fixed mean waveform from five early pulses in public SEEG data. We subtracted it from five later pulses. We divided the residual power by the sample variation within the later block. At 2 to 30 Hz, the mean paired change in this ratio was 0.16 dB from before pulses to 100 to 500 ms after pulses. Six of eight patients had a positive change. The patient bootstrap interval was 0.03 to 0.31 dB. The exact sign-change test gave p = 0.0625. This small sample does not give firm evidence of a population effect. An exact identity separates later block variation, early template noise, and a signed mean-shift term. The mean-shift interpretation requires errors with zero expected products across pulses and blocks. Contact order and time changed together. The package includes new cross-block products, native waveform examples, figures, source hashes, code, and a full retrieval recipe. It does not measure neural power or BCI task performance.

stereo-EEG · intracranial EEG · template subtraction · spectral power · brain-computer interfaces · measurement error · STReEF

Admitted 2026-11-09 18:45 UTC2 citing publications

References

usesFive-pulse estimates showed a greater mean 100 to 500 ms rise in repeatable power fraction at 2 to 30 Hz than at 70 to 150 Hz in eight STReEF SEEG runs sci:sha256:552984215d5744e6e94b5e864c21d391db78acd541661c7f6652caab5f35404d
We use the fixed SEEG sample, source metadata, pinned range hashes, source utilities, and Fourier power method. We retrieved all selected raw ranges again and added cross-block products for the fixed template audit. The patients and selection are shared.
usesA five-pulse check selected SEEG pairs with 60.8% mean shape agreement after contact order reversal at 20 to 500 ms in eight sessions sci:sha256:35cdaeabceeac5a3be97a54a264240386954c3bac2e4558f03b24d86d28e2bee
The original metadata selection method and public range reader come from this work. We preserve its three source utility files and license through the immediate spectral source publication. The local audit rebuilds the fixed selection with that method.
usesdoi:10.18112/openneuro.ds005448.v1.0.0
The analysis uses STReEF/OpenNeuro ds005448 version 1.0.0 for its fixed eight-patient SEEG sample and source recordings.
supportsdoi:10.1016/j.ynirp.2025.100274
Cited for the source study and dataset description of structural and effective connectivity in focal epilepsy.
supportsdoi:10.1007/s12021-022-09567-6
Cited for the clinical iEEG BIDS workflow used by the source authors.

Cited by

2 publications

Public catalog entries on Sciscape.

usesA five-pulse SEEG test did not show a clear increase in the 4–30 Hz persistence gap after stimulation in eight people
This report supplied the corrected group mean difference identity. We use the same mathematical idea and state its error assumptions. The new contribution is the tested data comparison, not a new identity.
usesFull five-pulse subtraction increased mean 70–110 Hz threshold crossings at 100–500 ms after stimulation in eight SEEG patients with fixed raw thresholds
Its fixed-template result supplied related evidence about calibration noise in a different STReEF sample at 2–30 Hz. It helped define the question. This study uses HAPwave data and tests 70–110 Hz threshold crossings. No STReEF data were used.