A 200 ms delay in a fixed voltage veto raised high-gamma crossing excess by 4.28 percentage points in 23 SEEG sleep records

We tested the effect of warning age on a fixed discharge veto. We used public sleep SEEG features from 25 people with epilepsy. The later test used 425 discharge labels and 23 people. The reader used 70 to 110 Hz power. The veto used 4 to 30 Hz voltage. A warning that was 200 ms old raised event-minus-background crossing excess by 4.28 percentage points. The 95 percent person-bootstrap interval was 2.91 to 5.71 points. It retained 74.9 percent of the current veto reduction in the first 1 second after a label. It retained only 38.3 percent in the first 300 ms. A stricter background check gave similar results in 22 people. The contribution includes analysis code, results, figures, source hashes, and a raw-data retrieval recipe; the source feature files are linked externally. It reuses a cited sci data product. It contains no intended command or stimulation test. The result does not establish BCI accuracy or safe stimulation.

SEEG · intracranial EEG · signal processing · epilepsy · interictal discharges · high-gamma power · voltage veto · warning delay · brain-computer interface