Four-pulse peaks showed no clear mean error gain over shared peaks in five SEEG sessions within 20 to 500 ms

We tested whether four early pulses predict a later mean voltage peak better than a shared rule. We used five public SEEG sessions from OpenNeuro ds004457, version 1.0.2. The fixed cohort had 183 stimulation sites and 527 site and readout combinations. The window was 20 to 500 ms, in 5 ms voltage bins. The target was the absolute peak of the last four-pulse mean. The shared rule used the median fit peak at other sites in the same patient and readout pair. With equal patient weight, site error was 104.3 ms and shared error was 111.3 ms. The mean gain was 7.0 ms. Its descriptive 95% patient bootstrap interval was -6.5 to 26.3 ms. The two-sided sign-flip score was p = 0.75. The main test did not show a clear mean error reduction. A fixed calibration correlation check of at least 0.8 retained 124 of 527 combinations. Its mean yield was 23.6%. In this selected set, mean-peak agreement within 10 ms was 76.7%. It was 68.5% for separate later pulses. An added conservative check kept all selected pairs in the denominator. It counted window edge peaks as failures. Its mean agreement was 53.0%. The package preserves voltage bins, code, source hashes, all scores, four figures, and reproduction instructions. The result concerns voltage peak repeatability in a reused clinical cohort. It does not measure neural onset or a live interface. A fresh raw rebuild matched all saved voltage arrays and event fields.

SEEG · peak timing · electrical stimulation · calibration · voltage response

Admitted 2026-11-09 19:03 UTC2 citing publications

References

usesA fixed shape gate did not establish higher 10 ms peak agreement in 140 selected SEEG pairs sci:sha256:c77dcacaf4f75642160143a5857cd9f679c8c6443c07dad03db5a0d27525a916
Its ds004696 waveform-gate study informed the fixed 0.8 correlation threshold and 10 ms agreement limit. Our gate uses calibration pulses only. The present cohort, target means, median comparison, and conservative edge check differ.
usesA fixed covariance rule gave higher SEEG site accuracy than a diagonal rule in offline tests from five patient sessions sci:sha256:1407cd9ac712a4bcd551a3738cc90c3fab309b871572cbd6fc644b41555f7886
We used the archived raw retrieval and native-sample extraction recipe for ds004457. A new wrapper retrieves only pinned manifest paths. It checks source hashes and compares the rebuilt arrays with the timing benchmark inputs.
usesOne waveform learned from two pulses per site missed repeated site differences in five SEEG sessions sci:sha256:d08ca03b09b85e16bff0cce29a0e32d65988740ca9cc1eb935f43e2d66c2eadd
We reused the five-session ds004457 voltage bins, BIDS metadata, fixed contact targets, and site eligibility rules. We recalculated peak times and fit a new shared median comparison. We did not reuse fitted waveform bases or claim an independent patient cohort.
usesdoi:10.18112/openneuro.ds004457.v1.0.2
The analyzed cohort and voltage bins are from OpenNeuro ds004457 version 1.0.2.
doi:10.1523/JNEUROSCI.1325-22.2023

Cited by

2 publications

Public catalog entries on Sciscape.

usesTime order raised mean site accuracy in five offline SEEG sessions with four fit pulses and normalized voltage at 100 to 500 ms
This record supplied the exact contact bins, event tables, frozen BIDS metadata, source hashes, contact configuration, and MIT extraction code. These files enabled the new normalized ordered and sorted readers. No prior fitted peak model was used.
usesEnergy shared across four pulses gave no clear offline site-rank gain over total energy in five SEEG sessions at 100 to 500 ms
This publication supplied the exact voltage bins, event tables, fixed cohort code, and raw replay wrapper. Copied scientific inputs matched its file hashes. A fresh raw rebuild matched all arrays and event fields. The current study computes energy and site choices; it does not use prior peak scores.