Wake and N3 coupling patterns agreed in voltage from 29 SEEG patients and BOLD from 16 records after site and distance controls

We tested whether cortical coupling patterns agree between wake and N3 sleep after site and distance effects are removed. We used separate public human SEEG and fMRI studies. The original gamma-power test failed in 29 patients. Its mean residual pattern correlation was -0.0083, with a 95% interval of -0.0551 to +0.0403. A literature-based voltage follow-up gave +0.3461, with interval +0.2451 to +0.4381. This electrical result was exploratory. We saved it before the fixed fMRI outcome calculation. In 16 fMRI records, merged into 15 conservative initials groups, mean agreement was +0.3084, with interval +0.2638 to +0.3520. Equal Fourier power gave +0.2547 in voltage and +0.2037 in BOLD. All four voltage and BOLD tests passed Holm correction across six original and equal-power tests. The largest corrected p was 0.000244. Mean-signal removal, a 50 mm distance rule, and value capping retained positive means. The gamma-power failure remains. This is shared evidence for a measured pattern property at different physical bands and spatial samples. It does not show matching anatomical edges or a common causal mechanism. Fixed common sources can also produce retained patterns. Clinical coverage, short records, unknown fMRI person links, and missing motion and upstream processing details limit inference. The package contains plans, code, source hashes, all person results, controls, figures, and repeat instructions. A source repeat matched every biological scalar, geometry row, inclusion decision, and first phase draw exactly.

human neuroscience · SEEG · fMRI · sleep · coupling patterns · signal readout · Fourier phase · reproducible analysis