Independent bipolar phase tests reject controls with independent contact phases on beta edges with a common contact in 16 public SEEG recordings

We tested two phase tests with data from the first 60 seconds of 16 public SEEG recordings in OpenNeuro dataset ds004100, version 1.1.3. We fixed contact signal magnitudes and assigned independent random phases to each contact. We then formed bipolar channels. On 953 beta edges with a common contact, the test with independent bipolar phases rejected 99.996% of these controls at a nominal 5% level. The test with independent contact phases rejected 5.00%. On 766 edges without a common contact, the rates were 5.21% and 5.51%. We averaged rates across participants. The result applies to this conditional phase model. It does not measure neural coupling. The package contains source records, contact magnitudes, results, code, and instructions. Checks passed, and repeat runs for two recordings gave equal arrays and table rows.

stereo-EEG · bipolar montage · phase randomization · magnitude-squared coherence · contact sharing · Monte Carlo test · conditional null model