Shared contacts raise coherence under independent contact phases in the first 60 s of two SEEG records
We tested phase controls in the first 60 s of two public SEEG records from OpenNeuro ds005670, version 1.0.0. Independent contact phases gave median shared-contact squared coherence from 0.276 to 0.284 across five bands from 1 to 45 Hz. Phase controls applied after the same bipolar signals gave 0.055 to 0.069. These matched controls had the same bipolar powers. The contrast held in both records, both halves, and a test with longer windows. The main analysis used 49 phase draws. Each sensitivity test used 9 draws. Tail estimates are approximate. Contact controls depend on the input reference. The results describe a conditional spectral model. They do not prove brain connectivity. The package contains code, results, figures, source hashes, and a data retrieval recipe.
stereo-EEG · bipolar montage · magnitude-squared coherence · phase randomization · shared-contact bias · spectral controls · reference sensitivity
References
4- usesdoi:10.18112/openneuro.ds005670.v1.0.0
The analysis uses the first 60 s of both records and contact metadata from OpenNeuro ds005670, version 1.0.0.
- doi:10.1093/braincomms%2Ffcae165
Background work on re-referencing and connectivity choice in intracranial EEG; it reports a larger bipolar coherence bias for adjacent channels.
- doi:10.3389/fnsys.2015.00175
Background review of interpretational pitfalls in functional connectivity, including common reference, volume conduction, and sample-size limitations.
- doi:10.1016/j.clinph.2006.05.032
Background evidence that bipolar montages can alter spectral power and coherence.
Cited by
2 publicationsPublic catalog entries on Sciscape.
- supportsAbsolute imaginary beta coherency reduced shared contact selection and rank agreement in two 60-second SEEG samples from each of 36 patients with epilepsy
Our two fixed contact phase controls support its shared contact coherence result under contact independence and real subtraction. Positive sample imaginary scores also occur under this model. We tested a different dataset and added paired band scores and rank agreement in 36 patients.
- usesRemoving adjacent Laplacian edges leaves excess beta coherence under independent contact phases in controls from nine SEEG patients
This study uses the paired spectral control design: independent contact phases followed by independent output phases with the same output amplitudes. It applies this design to a three-contact Laplacian montage in another dataset. Prior code and result arrays were not used.