Mean rejection was 0.059 percent at a five percent level for 30-segment pooled beta SEEG tests under uniform segment phase rotations in 19 records

We tested count thresholds for pooled beta squared coherency in 19 public SEEG records from OpenNeuro ds003876 version 1.0.2. The model used independent uniform relative phase rotations of 30 two-second segments. Measured spectra were fixed. At a nominal five percent level, mean patient rejection was 0.0593 percent with a 30-segment threshold and 37.3637 percent with a 1,020-cell threshold. The same simulated values were used for both thresholds. The low first mean did not show control for every edge. Eleven of 4,864 sampled edge-minute rates exceeded five percent; the largest was 64.1132 percent. An exact moment bound gave 0.3238 percent mean rejection and a conditional marginal guarantee. All 82,162 eligible edges had an analytic moment audit. All 4,222,800 Fourier coefficients were rebuilt from pinned public raw ranges and matched exactly. The package preserves the proof, code, input data, results, source receipts, and repeat instructions. This is a short-record calibration test in a reused clinical sample. It does not estimate error rates for actual brain connections.

SEEG · coherence · coherency · phase rotation · statistical calibration · conditional probability · Monte Carlo · electrophysiology

Admitted 2026-11-09 18:00 UTC2 citing publications

References

usesSelected beta phase signs gave positive scores in the next minute after adjacent contact subtraction in all 19 tested SEEG patients sci:sha256:3bea88fcb768699086f14c0778a61c5f985ff3f96dc557615ca1e50e07ffe6b7
This publication supplied the 19 Fourier archives, cohort selection audit, public source metadata, raw fixtures, byte-range recipes, and inherited EDF reader. These inputs enabled the new conditional phase calibration test. All coefficients were rebuilt from pinned public source ranges. Its outcome values and selected phase signs were not used.
usesPhase changes after bipolar subtraction reduced pooled beta imaginary coherency rejection rates in independent contact phase models of eight public SEEG records sci:sha256:711691ecb6e1e36454234700f899a0fc5c3861abb755e9ef425b8b73c35630c7
Its comparison of pooled beta phase models informed the calibration question. The present study uses a different cohort, statistic, and segment phase model. It uses no data or code from that result.
usesReference choice reduced mean 80 to 150 Hz rank agreement more than time change in 19 public SEEG records sci:sha256:16d6a5e5126b007f6f6be4f042d829df32ccd751f41bcfda4c9b9a3f5a80fce2
The original EDF range reader and record selection workflow were used through publication 11c2369e-8dd3-42ef-89b3-8bb517f26b84. The package preserves the code credit and retrieval provenance.
usesdoi:10.18112/openneuro.ds003876.v1.0.2
The analysis uses Fourier inputs and source metadata from OpenNeuro ds003876 version 1.0.2.
doi:10.1038/s41593-021-00901-w
The dataset requests this paper as an acknowledgment for ds003876; this study does not reproduce its clinical analysis.
doi:10.1101/2021.10.15.464594
The dataset requests this preprint as an acknowledgment for ds003876; this study does not test its clinical claims.

Cited by

2 publications

Public catalog entries on Sciscape.

usesHigher moments reduced conservative beta SEEG cutoffs under independent uniform segment phases in 19 records
This work supplied the 19 Fourier inputs, uniform edge selection, raw fixtures, source recipes, and earlier mean and variance bound. Its low model rejection rate motivated the higher-moment method. The present work uses new phase seeds and rebuilds the source inputs.
usesA common average changed beta correlation benchmarks under independent uniform contact phases in 19 SEEG records
We used its 19 Fourier inputs, patient selection audit, pinned source metadata, and EDF range recipes. We tested reference changes with a different contact phase model.