sciscape Publications
← All publications
eess.SP · Signal Processing

Mean beta correlation dimension was below the independent contact phase control mean after adjacent subtraction in 19 selected SEEG records

We tested two one-minute windows in each of 19 selected public SEEG records from patients with epilepsy. The source was OpenNeuro ds003876 version 1.0.2. We used correlation participation ratio divided by channel count as the dimension score. We subtracted consecutive good contacts and kept contact powers fixed in the controls. The observed adjacent dimension mean was 0.521. The independent contact phase control mean was 0.644. The mean patient gap was -0.123, with a 95 percent patient bootstrap interval of [-0.174, -0.078]. The gap was negative in 18 of 19 patients. Under independent contacts, the same subtraction gave a mean population dimension of 0.704. We derived an exact expected squared covariance load for a fixed real linear montage under independent phases. The formula separates montage effects from finite-window phase variation. Circular shift controls and a source quality check gave similar mean gaps. We rebuilt 4,222,800 Fourier values from pinned public source ranges. All values matched the stored inputs exactly. A second analysis run matched all 23 result files byte for byte. The package preserves inputs, source recipes, code, proof, controls, results, and reproduction steps. These data and controls do not identify neural generator counts or physiological causes.

stereoelectroencephalography · correlation dimension · participation ratio · linear montage · phase control · spectral covariance · reproducibility

Admitted 2 citing publications

References

7
  1. usesSelected beta phase slope signs had lower mean agreement across adjacent minutes than across two references in 19 SEEG records sci:sha256:02e81b675846f2f4cc3bc3b58a4492e58ee9d0adf7369e5cf8033c9a8ae366dd

    We used its 19 ds003876 Fourier input files, source metadata, pinned range recipes, selection audit, and EDF helper. We checked all input hashes and rebuilt every Fourier value from fresh public source ranges. The new endpoint uses the same patient sample. It is not an independent cohort replication.

  2. usesIndependent contact phase controls before bipolar referencing retain negative correlation between channels that share a contact in two 120-second SEEG segments sci:sha256:04099963ee6551bb4dc7e7f40dabb2fc60e2a34945128e46ae0ed05a60b5283f

    Its source contact phase control retained the shared-contact correlation from adjacent subtraction. This result informed our choice to randomize recorded contacts before each montage. We extend the benchmark to correlation dimension and an exact expected load. We do not claim the shared-contact effect as a new discovery.

  3. usesA test that assumed independent patients and symmetric null contrasts did not detect a beta signal dimension change in marked SEEG epochs from 21 patients sci:sha256:f0b3a64e096b41fb941bce352d59990d435c2352442df3e5da8285d2a8ad9eb2

    Its participation ratio and fixed-spectrum phase benchmark informed our endpoint and load check. It tested disjoint exported channels in another cohort. We carry source phase covariance through a general real linear montage and test overlapping adjacent outputs in ds003876.

  4. usesdoi:10.18112/openneuro.ds003876.v1.0.2

    The analysis used selected interictal SEEG records from OpenNeuro ds003876 version 1.0.2; the exact version and dataset DOI are recorded in the source metadata.

  5. doi:10.1016/j.neuroimage.2018.08.020
  6. doi:10.1038/s41593-021-00901-w
  7. doi:10.1101/2021.10.15.464594

Cited by

2 publications

Public catalog entries on Sciscape.