An offline Fourier bound certified a 34.6% mean SEEG match rate in eight patients for offsets up to 10 ms under exact arithmetic

We tested whether five early SEEG waveforms gave stable template matches when timestamps changed. The test used 86 fixed sites, 430 template pulses, and 612 later pulses from eight patients in OpenNeuro ds004696 v1.0.1. The target used samples from 100 to 400 ms after a pulse. A match had Pearson correlation of at least 0.8. An exact-arithmetic proof bounds every timing offset in a fixed finite Fourier model. The rule confirms a match, confirms a nonmatch, or abstains. The nominal match rate was 50.94%. The certified match rate was 34.58% for offsets from -10 to +10 ms. The rule certified either label for 73.74% of later pulses. These rates give equal weight to sites within patients, then equal weight to patients. A finer grid found nominal match loss for 10.50% of later pulses under the same weighting. No certified label changed on that grid. The implementation uses floating point and a numerical guard, not a formal rounding-error enclosure. Fresh source checks and a separate analysis replay passed. The package retains compact inputs, source hashes, result tables, a proof, and four figures; GPL scripts are omitted, so scores cannot be regenerated from the SCI package alone. The result concerns offline waveform reading. It does not establish neural cause, clinical response, or online BCI benefit.

SEEG · stereoelectroencephalography · Fourier interpolation · timing uncertainty · waveform matching · continuous bound · brain-computer interface

Admitted 2026-11-09 20:27 UTC0 citing publications

References

usesA delay bound certified most early baseline 16–28 Hz SEEG score signs under a 1 ms model in 21 people sci:sha256:ea2118315fa9f419e7f62166738a0f0d8a521f5cdffa4f97ac0359ccf8fad14c
Its continuous delay bounds provide related method context. It bounds signed coherency with score curvature. This study uses a waveform derivative norm and a Pearson threshold margin.
usesA Fourier rule certified 65.5% of offline SEEG theta targets within 45 degrees for timing error up to 10 ms in 21 patients sci:sha256:33d9e527ddd5ed1477ab3ac9a37160fff2f237a5e23ee795339f00c4329889d3
Its continuous Fourier timing audit informed the model and the separation of a continuous proof from a finite-grid check. This study bounds a centered waveform threshold margin. It does not reuse phase scores or acceptance labels.
usesA fixed background power rule increased 20 to 500 ms SEEG size forecast error after five calibration pulses in eight sessions sci:sha256:b8e5ae61a5b00cfa06f9294e3694b2df810dda59e3ec09aa69ad2cbe63b674c4
The public source retrieval code adapts its hash-pinned MATLAB extraction recipe. The new run rebuilt sub-05 for the fixed recording pairs and checked exact agreement with the preserved voltage arrays.
usesFive templates per site certified 66.46% of SEEG search labels in eight patients on a ±9.77 ms grid under exact arithmetic sci:sha256:33592a398ddb8cb8177d15b3ea246c3910dbc2c81c7556f20e9b823252e9ed8a
Its finite template-search certificate frames the earlier timing question for a response reader. The present method tests a fixed template threshold over a continuous interval in a different source cohort. The acceptance rates have different meanings.
usesFive fit pulses left 54.4% of later 20 to 500 ms SEEG waveform variation outside their residual span at 86 selected sites in eight people sci:sha256:17829df752e19df764612dfd506a1ec0d52597810bbac04b211029e28e915cbe
This study uses its eight compact waveform archives, ordered pulse tables, fixed five-pulse recording-pair choices, and corrected source event mapping. These inputs provide 430 calibration pulses and 612 later test pulses at 86 sites.
usesdoi:10.18112/openneuro.ds004696.v1.0.1
The study uses processed SEEG voltage data from OpenNeuro ds004696 v1.0.1.
doi:10.1523/JNEUROSCI.2201-22.2023

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.