Separate pulse groups and a control before each pulse gave positive difference estimates for successive blocks with opposite contact orders in three of eight STReEF SEEG patients

This study tested remote waveforms from successive stimulation blocks with opposite contact orders. It used public STReEF SEEG data from OpenNeuro ds005448 version 1.0.0. A fixed plan selected 96 block pairs and 960 pulses from eight patients. Separate pulse groups gave a squared-difference estimate under independent zero-mean errors. The primary endpoint subtracted the same measure before a pulse. Three of eight patient means were positive. The group mean was 7,106 square microvolts. Its patient bootstrap 95% interval was -1,100 to 21,746. The exact two-sided sign-flip p value was 0.469 under symmetric patient errors. Removing STREEF07 reduced the mean to 52. The data did not establish a group difference or agreement between orders. Fixed block order, pulse history, and possible artefact prevent a causal polarity claim. The package supplies paired epochs, source metadata, exact object versions and range hashes, code, results, and instructions. Fresh checks matched all 960 source ranges and processed arrays. An independent calculation reproduced the primary result. A proof and simulation show when separate pulse groups remove squared-noise bias and when shared errors remain.

stereo-EEG · single-pulse electrical stimulation · contact order · cross-validated distance · stimulation artefact · epilepsy · paired waveform analysis · OpenNeuro