A scale from five early pulses reduced full template SEEG error at 20 to 500 ms after contact order reversal in eight people

Can five early pulses set a response scale before any later pulse is used for fitting? We tested a bounded scale from early pulse variation in public STReEF SEEG data, ds005448 version 1.0.0. The fixed sample had eight people with focal epilepsy, 96 conditions, 960 pulses, and 6,803 readout-condition rows. At 20 to 500 ms after recorded contact order reversal, the scale reduced normalized squared error by 9.78 percent against the full early mean. All eight people had lower error. The Holm p value was 0.015625. Six people had lower error than zero, but the planned exact test did not establish benefit over zero (p = 0.0625). Before the pulse, zero had lower error than the scale in all eight people. The result is exploratory in a reused clinical sample. Contact order and time change together. It does not establish a clinical or online interface benefit. The package contains formulas, code, frozen inputs, three figures, source hashes, and a tested public retrieval recipe.

stereo-electroencephalography · SEEG · brain stimulation · response forecasting · signal scaling · focal epilepsy · brain-computer interface

Admitted 2026-11-09 18:18 UTC10 citing publications

References

usesSigned SEEG alignment at 20 to 500 ms exceeded a before-pulse control in eight people after contact order reversal sci:sha256:566477652590c53f83112c199b70fdca7b0e1f47861e04e334107dda534bae03
This study uses the frozen pulse products, selection, source receipts, and source checking code from this report. Its full-template result was known before the local plan and motivated the new early pulse scale.
usesA scale fit from one pulse lowered mean SEEG forecast error in six of eight sessions after contact order reversal at 20 to 500 ms sci:sha256:18c6ed2417c8fa21e9a22c8580a0e0bf89683a8de79f7dbda88c0cbd316fe6c2
The fixed source sample and retained earlier code trace through this publication. Its new-pulse scale result was known. This study tests a scale from five early pulses without its response gate or later pulse fitting.
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
This publication supplied the original metadata selection and extraction for the fixed STReEF sample. This package retains that source code and provenance through the direct input publication.
usesdoi:10.18112/openneuro.ds005448.v1.0.0
The fixed analyzed source sample is STReEF, OpenNeuro ds005448 version 1.0.0.
doi:10.1016/j.ynirp.2025.100274
doi:10.1007/s12021-022-09567-6

Cited by

10 publications

Public catalog entries on Sciscape.

usesTwo early pulses gave 3.3 percent higher mean relative normalized SEEG forecast error than five after contact order reversal in eight people at 20–500 ms
Used its fixed eight-person STReEF sample, exact waveform products, source provenance, and retrieval code. Rebuilt all 960 public pulse ranges and tested a new fixed two-versus-five calibration budget.
usesWith four fit pulses, mean SEEG shape change was 1.04 percentage points above a before-pulse change at 20–500 ms in eight people
Its scale test in the same cohort informed the research scope. We instead test a completed-window shape fraction with a separate early reference and a matched before-pulse change. Its result scores are not inputs.
Early repeatable readouts accounted for 74 percent of mean SEEG error reduction against zero at 20 to 500 ms in eight patients after contact order reversal
[Legacy relation: reproduces] The new raw rebuild and analysis reproduce four practical forecast losses in four windows for all 96 conditions. The largest absolute condition loss difference is 3.14e-11. The same sample supplies no new patient evidence. The added check partitions gains by fixed early repeatability.
usesSigned templates had mean AUC 0.7925 versus 0.83625 for energy at 20–500 ms after contact order reversal in eight SEEG sessions
The source byte-range extraction code is reused through the cited parent archive. A small change returns all candidate native voltage differences. The source hash checks, channel checks, and voltage scale checks remain in use.
usesA joint test with five later pulses retained 449 of 1,563 SEEG calls at 20 to 500 ms after contact order reversal in eight people
This study retains its fixed 96-condition sample, all 6803 SEEG readout-condition inputs, native audit waveforms, metadata, and pinned source retrieval code. It adds joint signed response tests; it does not reuse the parent amplitude result as new evidence.
usesMixed five pulse tests gave lower SEEG waveform error than contact order transfer in eight people
This study uses its fixed SEEG sample, native pulse products, metadata, and pinned source retrieval code. It adds a complete split test for the full five pulse mean reader.
usesA joint SEEG check with five early pulses accepted 71.0 percent of later 20 to 500 ms maps after contact order reversal in eight people
The study uses the frozen 96-condition STReEF sample, all 6803 readout-condition rows, exact native waveform Gram matrices, source metadata, pinned range records, and extraction code. A fresh retrieval reproduced all 960 ranges and all compact inputs exactly.
usesA rule that used half of one SEEG monitor's gain change did not establish pulse tracking after contact order reversal in eight patients
Used its MIT metadata selection and range retrieval helpers through the direct input. Adapted its early variation scale for an added reference on the same 16-mode targets. This reference was added after the primary result.
usesFixed donor variance weights raised mean SEEG pulse AUC with 16 temporal modes after five fit pulses in eight people from a separate cohort
Uses the fixed 96-condition, 6,803-readout target sample, raw range receipts, source metadata, audit waveforms, and retrieval helper. New source reads verify all 960 ranges before an independent donor-weight comparison.
usesSeparate scales for 16 slow modes increased mean relative SEEG forecast error at 20–500 ms in eight patients after contact order reversal
We adapted its early pulse variance rule to one shared condition scale and separate mode scales. Its one-scalar-per-readout rule is also retained as a reference on our projected target. Its published native-voltage scores are not directly compared with these scores.