A joint test with five later pulses retained 449 of 1,563 SEEG calls at 20 to 500 ms after contact order reversal in eight people

A joint test retained 449 of the 1,563 SEEG readout calls made by separate tests. The study used eight people, 96 fixed stimulation conditions, and five early plus five later pulses per condition. The later pulses had reversed recorded stimulation contact order. The window was 20 to 500 ms. The test used all 32 shared pulse sign patterns and a maximum statistic across readouts. It made five before-pulse calls, compared with 174 from separate tests. The equal person mean joint post-minus-before difference was 6.222 percentage points. The exact person sign-flip p value was 0.015625, under independent people and symmetric null differences. Joint error control requires sign invariance of the true-null pulse scores and applies within one condition and window. The data do not establish this assumption or a neural cause. Five pulses cannot support a two-sided exact call at 0.05. The package contains complete inputs, code, figures, proof, source receipts, and repeat instructions. All 960 pinned pulse ranges were checked again. All reconstructed inputs matched exactly.

stereoelectroencephalography · SEEG · intracranial EEG · sign-flip test · maximum-statistic correction · pulse responses · epilepsy