One site with nearly constant signals supplied 90 percent of the corrected 20 to 100 ms contact-order difference estimate in eight STReEF SEEG patients

We compared two stimulation contact orders at 96 sites in all eight SEEG patients in STReEF. Each site had five pulses per order. The public source was OpenNeuro ds005448, release 1.0.0. Separate pulse repeats estimated measured waveform difference energy from 20 to 100 ms. We subtracted a matched control before stimulation. The equal-patient mean was 11.7298 baseline variance units. The planned exact two-sided patient test gave p = 0.0625. One site supplied 89.93 percent of the net estimate. At this site, two native contacts had large steps and nearly constant post-pulse values. The source marked both contacts good. Omitting the site in a later audit lowered the mean to 1.1920. This finding concerns signal quality. It does not identify a neural polarity effect. Contact order and elapsed time varied together. The package contains all 960 processed pulse epochs, native examples, source versions and range hashes, an energy identity proof, code, results, and repeat instructions. The primary analysis remains unchanged. The signal audit and removal rules were chosen after the primary result.

stereoelectroencephalography · SEEG · contact order · stimulation artifact · signal quality · difference energy · OpenNeuro · STReEF