Centered averaging let later voltage change computed gamma power before onset with causal filtering in 64 SEEG records

We tested whether power before an onset mark depends on later voltage. We used 64 public SEEG seizure records from 32 patients in OpenNeuro ds004100 version 1.1.3. Each patient had eight fixed adjacent contact pairs and two runs. We kept all voltage before each mark fixed. We replaced later voltage with the last earlier sample. We used a causal 70 to 110 Hz filter and a centered 200 ms power average. The mean patient absolute change was 0.612 dB. The descriptive 95 percent patient bootstrap interval was 0.554 to 0.670 dB. The same causal filter with trailing averaging gave exactly zero change. An earlier cut gave a similar result. The test did not show an effect specific to seizure onset. A proof gives the average's possible time dependence. Independent calculations, an offline repeat, and a second source download passed. The package preserves native digital samples, code, provenance, results, and repeat instructions. These results concern the tested processing method and artificial voltage changes. They do not establish neural activity before onset or clinical value.

signal processing · intracranial EEG · SEEG · seizure · gamma power · causal filtering · centered averaging · time dependence · reproducibility