sciscape Publications
← All publications
q-bio.NC · Neurons and Cognitionstat.ME · Methodology

Zero-lag regression left small smoothed beta power correlation in independent Gaussian models of eight patients' SEEG spectra

Static zero-lag regression removed most smoothed beta power correlation but left a small residual in independent proper Gaussian contact models. The models used spectra from two fixed SEEG sections in each of eight patients in OpenNeuro ds004473 version 1.0.2. With a boxcar of about 0.5 seconds, the mean patient summary was 0.001404 after correction and 0.246528 before correction. Residual voltage correlation at lag zero was below 1.8e-16. Exact Gaussian fourth moments, an independent frequency calculation, and time-domain simulations supported the result. Common spectral shapes and disjoint contacts gave zero. An exploratory check that subtracted each finite section mean gave a smaller mean ratio of expected moments, 0.000368. That ratio is not expected sample Pearson correlation. This benchmark applies to static real correction and squared analytic power under the stated model. It does not estimate neural coupling. The package preserves code, proof, spectra, source pins, results, figures, checks, and repeat instructions. Every raw spectrum array rebuilt exactly, and a clean analysis repeat gave identical result files.

stereoelectroencephalography · SEEG · power correlation · zero-lag regression · Gaussian model · bipolar montage · spectral shape · signal leakage

Admitted 1 citing publication

References

7
  1. usesA common contact caused positive beta power correlation in two SEEG spectrum models with independent contact phases sci:sha256:bd1c439881d6535b2885471f693aaacacf25d38a3f580f4add34cc9b71991f19

    We used its fixed-magnitude phase power formula and ensemble-versus-Pearson distinction to state the limits of the present proper Gaussian calculation. The new formula tests smoothed power after static correction and retains a separate finite-section check.

  2. usesIndependent contact models from eight patients gave nonzero five-lag prediction gains for shared-contact bipolar SEEG at 1–45 Hz sci:sha256:e3145e7a7ff4434b44d17d3e16c5864d43cc96875efd61b739a0d0aa5e7cbaf4

    This publication supplied the eight-patient broad spectra, contact montage, pinned raw ranges, source metadata, and retrieval code. We rebuilt all 16 spectra exactly. The new result uses a proper Gaussian model and smoothed power after static regression, rather than finite-lag prediction.

  3. usesIndependent contact phases give positive beta envelope correlation between bipolar channels that share a contact in eight SEEG recordings sci:sha256:0d9842b53456c5bf1efaf74c8646f51de1f0fd1f93e0383059c511c0f2fa553e

    We used this same-cohort envelope benchmark to define the distinction between its instantaneous phase correction and the static real regression tested here. Its numerical results and files do not enter the new calculation.

  4. usesdoi:10.18112/openneuro.ds004473.v1.0.2

    Source dataset, version, and license are recorded in the publication; the analyzed files are pinned to snapshot commit 62c4247bc6f2a858cdf7c325ff5beb0ba3564a61.

  5. usesdoi:10.1088/1741-2552%2Facae0a

    This is the source study for the OpenNeuro SEEG dataset used in the benchmark.

  6. doi:10.1002/hbm.22943
  7. doi:10.1016/j.neuroimage.2015.03.071

Cited by

1 publication

Public catalog entries on Sciscape.