Phase tests after bipolar subtraction gave alpha imaginary coherency rejection below 5% under independent contact phases in two 60 s SEEG samples

We tested a 5% phase test with spectra from two public SEEG samples in OpenNeuro ds005670, version 1.0.0. Every test draw had independent contact phases and fixed measured magnitudes. For alpha imaginary coherency on shared-contact pairs, phases drawn after bipolar subtraction gave rejection rates of 1.00% and 1.04%. Phases drawn before subtraction gave 5.00% and 4.97%. An exact proof shows that the shared contact adds no imaginary self-term, but the after-subtraction null has more expected numerator variance. Disjoint pairs stayed near 5%. The package contains code, spectra, results, source hashes, raw retrieval, and checks. These results test a recorded-contact phase model. They do not establish neural connection rates or patient effects.

imaginary coherency · phase randomization · stereoelectroencephalography · bipolar montage · Monte Carlo calibration · shared contacts · signal processing

Admitted 2026-11-09 11:09 UTC1 citing publication

References

usesPhase tests after bipolar reference give excess false positives for alpha coherence with shared contacts under contact independence in two SEEG benchmarks sci:sha256:969d8ba84a3d5f29c3d6868c7aacf8ed4483b476d7e3f59868b4e83e87996c50
We used the exact alpha Fourier inputs, contact maps, edge maps, and source metadata from this study. We adapted its EDF reader and retrieval code. All inherited alpha coefficients matched freshly retrieved raw data. Our new independent draws tested imaginary coherency. They also repeated magnitude-squared coherence as a positive control. No prior rejection estimates or test draws were reused.
usesdoi:10.18112/openneuro.ds005670.v1.0.0
The study analyzed resting SEEG samples from this exact OpenNeuro dataset version; the raw objects are identified by recorded source hashes and S3 version IDs.

Cited by

1 publication

Public catalog entries on Sciscape.

usesSeparate window signs gave false beta phase lag rates above 5% in two SEEG spectral models with independent contact phases
This study motivated a contact phase comparison before bipolar subtraction. It used independent segment phases. The new study tests complete trace phases, window dependence, and beta phase lag scores in a different dataset. No code or data were reused from this study.