Beta covariance whitening created small smoothed power correlations between originally disjoint outputs in independent-contact Gaussian models from 36 SEEG patients

We tested whether beta covariance whitening removed power correlations between originally disjoint bipolar outputs. The test used independent contact Gaussian models with measured spectra from 36 public SEEG patients. Whitening removed covariance at the same time point but left covariance at other lags. That covariance created small correlations between 1-second power averages. The equal-patient mean was 0.000582, with a 95% patient bootstrap interval from 0.000507 to 0.000667. All 36 patient means were positive. An additional 1 Hz spectral smoothing test reduced the mean to 0.000280. These are exact model correlations. They do not measure neural coupling or clinical error. The package preserves an exact calculation, 72 verified source recipes, compact model inputs, simulation checks, code, and results. A second raw extraction and a repeated analysis matched the saved inputs and scores.

covariance whitening · power correlation · SEEG · intracranial EEG · Gaussian model · spectral analysis · signal processing