An average-to-Cz reference mixture reversed the ON-minus-OFF 13–30 Hz C3/C4 squared imaginary coherency contrast between positive endpoints in ds002778 v1.0.5 sub-pd13 with fixed windows
We made a tool to bound squared imaginary coherency over constant mixtures of the 32-channel average and Cz EEG references. The same mixture weight was used in both visits. We tested C3/C4 at 13 through 30 Hz in OpenNeuro ds002778 v1.0.5. A fixed window screen retained six ON/OFF pairs. In sub-pd13, both endpoint contrasts were positive. A mixture with Cz weight 0.1875 gave a negative contrast of -0.00155802. Both signs occurred in four pairs; endpoint tests detected them in three. The method gives outer bounds and attained point enclosures at tolerance 1e-6 for a fixed coefficient model. All six planned settings, a 1e-7 check, independent numerical checks, code, source hashes, Fourier data, and both raw sub-pd13 records are preserved. Results depend on the window and band rules. They do not establish a medication effect or a direct brain connection.
EEG · imaginary coherency · reference choice · spectral analysis · interval bounds · Parkinson disease · C3/C4
References
4- usesAverage and Cz references gave opposite estimated C3/C4 squared imaginary beta coherency contrast signs in 3 of 6 accepted ds002778 pairs sci:sha256:d2a0ce78e591983f65f27e207940d8c29630251e336607f6e42c448891913c2f
We reused its 30 Fourier files, window ranges, source records, and BDF extraction code. They enabled the full reference mixture test.
- doi:10.1088/1741-2560%2F13%2F3%2F036016
- doi:10.1016/j.clinph.2004.04.029
- usesdoi:10.18112/openneuro.ds002778.v1.0.5
The analysis uses the version-pinned ds002778 v1.0.5 EEG records; source hashes and URLs are preserved in the archive.
Cited by
1 publicationPublic catalog entries on Sciscape.
- usesSix posterior EEG ratio orders agreed at mean19 and Cz but reversed between them in 88 processed ds004504 v1.0.9 records
Read after the main analysis as prior evidence that reference endpoint agreement can miss an interior contrast sign. It supplied no numerical input. The present evidence concerns person order and group AUC for a different measure and cohort.