Theta trigger averages exceeded time-shift controls and forecast correction had an uncertain effect in seven SEEG rest recordings without pulses

We tested a no-pulse control for brain interfaces that use a rhythm to time stimulation. We used 28 selected SEEG pairs from seven public rest recordings in OpenNeuro ds006065 version 1.0.0. The trigger used only past voltage. Fit, calibration, and test blocks were separate. At 20–300 ms, mean trigger-average RMS was 0.2438 fit standard deviations. Circular time-shift controls gave 0.1182. This planned secondary difference was positive in all seven people. The main forecast correction result was uncertain. Mean RMS decrease was 0.0505, with a descriptive patient bootstrap interval of −0.1463 to 0.1988. The exact patient sign-flip value was 0.640625. Correction reduced the average in five people and increased it in two. The forecast had no clear advantage over an earlier trigger-average template. A future-only additive response was preserved under fixed triggers and a fixed forecast. This code check does not establish recovery of a real brain response. The package includes exact voltage, fresh full-source hash checks, code, models, control epochs, results, figures, and repeat instructions. The study supplies a reproducible SEEG benchmark for a known phase-selection problem. It does not test actual stimulation effects or clinical benefit.

SEEG · theta triggers · no-pulse control · forecast correction · time-shift control · phase selection · ridge regression