Time-centered 50 to 250 ms SEEG windows showed no clear increase in early-to-late change after stimulation in five patients

We tested whether early and late pulse means differ more after electrical stimulation than before it. The public SEEG data came from five patients in OpenNeuro ds004457 version 1.0.2. We retained 150 ten-pulse blocks and 1,067 site–receiver links. Each 50 to 250 ms waveform had its time mean removed. We divided the first-five versus last-five squared distance by its exact mean over 126 balanced splits. The reference ratio has mean 1 and range 0 to 9 when pulse scatter is positive. Equal-patient mean ratios were 0.999683 after stimulation and 1.041229 before it. The contrast was -0.041545. The joint split reference gave p=0.24731. This test assumes uniform balanced allocations within blocks and independent allocations between blocks. The exact patient sign reference gave p=0.25. It assumes independent patient effects and null sign symmetry. The tests did not show a clear increase. They do not prove equivalence. All four planned checks are reported. Keeping window voltage levels changed the contrast sign. The package includes the proof, code, pulse windows, hashes, all reference values, actual splits, and decode instructions. A fresh native decode matched all 3,477,760 voltage bins. An offline repeat matched all 11 result files exactly. This analysis reuses a five-patient clinical sample.

stereoelectroencephalography · electrical stimulation · pulse order · waveform distance · exact reference · balanced splits · reproducibility · OpenNeuro