Simulated responses at three theta RMS changed six-cycle 7 Hz wavelet phase estimates 20 ms before events in 18 fixed SEEG rest clips

We tested whether a future signal can change a phase estimate before an event. We used 18 fixed 96-second SEEG rest clips from seven patients in OpenNeuro ds006065, version 1.0.0. The main test added a damped 7 Hz cosine with an amplitude of three baseline theta RMS. The signal began 20 ms after each known event. A symmetric 7 Hz Morlet wavelet with six cycles measured phase 20 ms before the event. The equal-patient mean rate of phase changes above 30 degrees was 27.24%. The 95% patient bootstrap interval was 26.03% to 28.57%. Mean pairwise phase consistency increased from -0.0049 to 0.1117. A fit that used only earlier samples had exactly zero phasor change in 2,124 checks. The results measure future dependency under the stated signal assumptions. They do not show a brain change before an event or a treatment benefit. The package contains public clips, fixed source records, simulated signal code, event estimates, a reusable dependency audit, figures, and repeat instructions.

SEEG · intracranial EEG · theta phase · wavelet · future dependency · phase consistency · pairwise phase consistency · stimulation artifact · signal processing