A sleep rise in joint magnitude beyond a phase reference was not confirmed in 50 SEEG patients and 16 fMRI records

We tested whether deep sleep raises joint signal magnitude above a reference that preserves full cross-spectra. Separate studies supplied 50 SEEG patients and 16 fMRI records. We grouped the fMRI records into 15 possible people by initials to reduce repeat counts. Each window became normal scores. We compared mean squared signal products with an exact finite-spectrum phase mean. N3 minus wake excess was -0.002016 in SEEG (95% person interval -0.012562 to +0.008469) and -0.030057 in fMRI (-0.077025 to +0.012561). Both primary Holm p values were 1. Excess in N3 was also uncertain in each modality. The fixed rule for a shared result failed. Several fMRI controls changed the score sign or size; none established the shared claim. Simulations show that normal ranks can shift the mean and common changing variance can create positive excess. Thus this score does not identify a neural mechanism. We provide code, an exact reference implementation, both cohort tests, all controls, provenance, figures and repeat instructions. Short records, unequal spatial samples, epilepsy, uncertain fMRI identities and missing nuisance traces limit inference. Source waveforms stay external under their terms.

deep sleep · SEEG · fMRI · Fourier phase · fourth moment · cross-modal test