Sample rank forced 92.8 percent of unrestricted complex canonical beta coherence scores to one in selected 20-second SEEG samples from 19 patients
We tested unrestricted complex canonical squared coherence in selected interictal SEEG records from 19 patients in OpenNeuro ds003876 version 1.0.2. The estimator fitted separate complex channel weights at each beta frequency. In 20-second samples, the sum of the two signal ranks exceeded the ten spectral observations for 931 of 1,003 contact group pairs. This rank condition forced 31,654 of 34,102 scores to one. All 20 independent contact phase controls gave the same count. The mean fitted score was 0.99369. Fixed weights gave a mean score of 0.04036 on a separate minute. These means gave each patient equal weight. A fixed three-channel reduction removed the forced values at this record length. Fitting bias remained below the rank limit. The rank limit is established linear algebra; the new result is its measured extent in these public records. The package preserves a rank-check tool, the analysis plan, source code, Fourier inputs, primary weights, source pins, results, and validation receipts. Original recording bytes reproduced all 4,222,800 archived Fourier coefficients exactly. The result describes this selected sample and estimator. It does not establish brain connections or clinical performance.
canonical coherence · sample rank · SEEG · intracranial EEG · spectral connectivity · finite-sample bias
References
6- usesThe common-reference-invariant beta component contributed more to agreement in 19 adjacent-minute SEEG records sci:sha256:f0d64ceacebda1587fbb8cf20d8bc531c139cbd98608e6166f45e2a4738cfa1d
Reused the 19 Fourier archives, raw examples, source pins, and contact quality records. Adapted its retrieval code. Computed new coherence, rank, phase control, and fixed-weight test results.
- usesReference choice reduced mean 80 to 150 Hz rank agreement more than time change in 19 public SEEG records sci:sha256:16d6a5e5126b007f6f6be4f042d829df32ccd751f41bcfda4c9b9a3f5a80fce2
Used the adapted EDF decoding helper through publication 8ebee4f3-5ebf-49b7-a80e-e7b62f44bb6b. This code supports byte-range decoding, physical scaling, contact mapping, and source time checks.
- usesdoi:10.18112/openneuro.ds003876.v1.0.2
This exact OpenNeuro release supplied the selected SEEG recordings used in the study.
- doi:10.1038/s41593-021-00901-w
- doi:10.1101/2021.10.15.464594
- doi:10.1016/j.neuroimage.2019.116009
Cited by
10 publicationsPublic catalog entries on Sciscape.
- usesA ridge penalty of one gave no clear beta coherence gain above a phase control after one minute of training in 19 SEEG records
This rank audit supplied the 19 Fourier archives, selection rules, quality checks, byte recipes, and retrieval code. We used them for the fixed ridge comparison. We also reproduced its 60-second unrestricted test scores.
- usesEqual name group weights changed fewer top beta strength nodes than a change of one minute in 19 selected SEEG records
Reused all 19 Fourier archives, source pins, metadata, contact quality records, raw examples, and retrieval code. Computed new beta edge strengths, weighting comparisons, and selected node sets. These are the same patients.
- usesSample rank forced unit imaginary coherency in unregularized 20 second beta SEEG fits with real weights from 19 selected patients
We use the 19 public Fourier inputs, metadata selection, contact group rules, source provenance, and retrieval code. We extend the sample rank audit to imaginary coherency with real spatial weights. This is an extension in the same data sample.
- usesFour-step bipolar pairs had higher mean 30–80 Hz exponents than one-step pairs in 19 selected SEEG records when 58–62 Hz was omitted
This work supplied the fixed 19-record cohort, source metadata, versioned EDF byte ranges and hashes, quality audit, and EDF helper code. We retrieved and verified the ranges again. We computed new matched-span spectra and new reference checks.
- usesFixed 50 percent shrinkage raised the mean later beta coupling score over a single bipolar pair after a 20 second fit in 19 selected SEEG patients
We used its complete 19 patient Fourier inputs, quality checks, source provenance, and public retrieval code. Its sample rank result motivated this fixed shrinkage comparison. We calculated new scores on a separate minute with frozen weights and phase.
- usesSelected covariance between named shafts reduced mean Gaussian beta loss for the next minute in 19 SEEG records
Its fitting-bias result informed the separate test minute and regularization. Its quality notes informed the exploratory contact checks made after the primary result. We used a different estimator and did not copy its model scores.
- usesAdjacent contact differences could not determine 44 percent of common-average beta power on average in 19 selected SEEG records
Its archive supplied selected records, pinned EDF byte ranges, source headers and metadata, and the EDF helper. We adapted its retrieval and metadata selection code. These files enabled a new montage power audit on the same 19 patients.
- usesA fixed ridge penalty increased beta squared coherence transfer in 9 of 19 selected SEEG records
Reused 19 Fourier inputs, raw examples, source records, contact quality records, group rules, and retrieval code. Calculated new ridge transfer and control results. Reproduced all 34,102 unrestricted 60-second cells.
- usesA unit covariance load reduced beta excess coherence in the next minute in 12 of 19 selected SEEG records
This study reused contact quality checks, montage rules, and the observation-space solver. It reproduced the unrestricted 60-second baseline and tested a new fixed covariance load on a separate minute.
- usesEach of 19 tested SEEG records had a pooled beta coherence leader change after fourfold group copies with fixed signals
Used all 19 Fourier archives, contact quality records, raw fixtures, source metadata, and retrieval code. Computed new group-weight and copy-test results. Fresh public source ranges rebuilt all saved coefficients exactly.