A 30 Hz template derivative captured more later SEEG variation after pulses than before pulses at 86 selected sites in eight patients

We tested a small waveform model for feedback after brain stimulation. The source was public SEEG data from OpenNeuro ds004696 version 1.0.1. We reused 86 selected sites in eight patients with epilepsy. Five early pulses defined a mean response with a 30 Hz cutoff and its orthogonal time derivative. Five separate later pulses tested the model at 20 to 500 ms. The derivative captured 6.18 percent of later centered variation after a pulse and 4.28 percent before a pulse. The mean difference was 1.90 percentage points. All eight patient differences were positive. The descriptive 95 percent patient bootstrap interval was 1.52 to 2.28 points. The two-sided exact patient sign-flip p value was 0.0078125 under independent-patient and symmetric-null assumptions. The amplitude direction and derivative together captured 24.2 percent after pulses and 23.0 percent before pulses. Thus, 75.8 percent of post-pulse variation was outside this model. Fixed frequency controls supported a small derivative advantage. The result describes waveform geometry. It does not identify neural timing changes or show a forecast or device benefit. The package includes exact voltage, pulse maps, code, fitted directions, scores, three figures, licenses, and a source retrieval recipe. Independent calculations and a full numerical repeat passed. A fresh source check reproduced one patient exactly.

SEEG · electrical stimulation · evoked potentials · waveform variation · template projection · time derivative · brain-computer interface

Admitted 2026-11-09 20:00 UTC3 citing publications

References

usesEight donor shapes gave less forecast error and 0.002 lower mean response AUC than eight cosines after five fit pulses at 86 SEEG sites in eight sessions sci:sha256:2a6f7f03fdc0a3d478dcf18e72daf6d41c9857ca607df84cba991a3aaed7402c
We used its exact compact voltage for all eight patients, pulse maps, fixed readout choices, and source provenance. These files supply every new numerical result.
usesFive calibration pulses gave mean later signed template AUC of 0.939 compared with 0.859 for one pulse in eight SEEG sessions sci:sha256:1d2aa103275ad7400befec4c899740ede8f2eb59a09897bdc3f4d4e190e54418
We kept its five-early-pulse selection of one remote bipolar readout at each of 86 stimulation sites. Its selection code and choice tables came through the direct compact input. We did not select contacts again.
usesTwo components from five early pulses captured 29.8 percent of later SEEG variation at 20 to 500 ms after removal of the early mean direction in eight patients sci:sha256:40710d623f4d60ce0573eff904e924868aabf89d82baccd4f1ea7b426e43eded
Its early residual-space test motivated a test of a specific amplitude and derivative basis. The readout sets, fitted bases, and denominators differ. We did not use its numerical scores for our estimates or a direct method comparison.
usesA fixed background power rule increased 20 to 500 ms SEEG size forecast error after five calibration pulses in eight sessions sci:sha256:b8e5ae61a5b00cfa06f9294e3694b2df810dda59e3ec09aa69ad2cbe63b674c4
We adapted its pinned full-source download and selected-voltage reconstruction recipe through the direct compact input. A fresh patient 05 download reproduced that compact input exactly.
usesdoi:10.18112/openneuro.ds004696.v1.0.1
The analyzed compact voltage and source BIDS metadata are from OpenNeuro ds004696 version 1.0.1; the recorded dataset license is CC0.
doi:10.1523/JNEUROSCI.2201-22.2023
doi:10.1371/journal.pcbi.1011105
doi:10.1016/j.heliyon.2023.e18671

Cited by

3 publications

Public catalog entries on Sciscape.

usesA 7 Hz phase input increased mean forecast error for later log response size after five calibration pulses at 86 SEEG readouts in eight patients
Used its fixed 86 readouts, response and earlier control voltage, event maps, source records, and GPL retrieval code. Added past voltage and a five-pulse forecast test.
usesEarly SEEG fits with a 9.77 ms shift limit increased later forecast error by 5.82% in eight patients
The template derivative result motivated a test of exact bounded shifts. The present test fits finite template shifts and compares known-target fits with early forecasts. It does not treat derivative projection as a neural delay or use the prior numerical scores as new evidence.
usesA five-pulse derivative update gave a 0.155% mean SEEG forecast gain at 86 selected sites in eight patients
This study used its compact source voltage, pulse maps, extraction code, and 30 Hz template derivative method. It added an early-to-late forecast test with calibration-only model fitting.