A fixed 30 mm test of 1 s SEEG windows did not establish greater repeated beta correlation change near motor stimulation in 15 patients

We tested whether repeated beta correlation change was greater near a motor stimulation site. We used all 35 released SEEG records from 21 patients in OpenNeuro ds006519, version 1.0.0. The fixed spatial rule required complete contact locations and both distance classes in each record. It retained 22 records from 15 patients. Near edges had at least one bipolar pair midpoint less than 30 mm from the stimulation pair midpoint. Far edges had both pair midpoints at least 30 mm away. We used two 1 s windows per labeled state, filtered separately at the native sample rate from 13 to 30 Hz. A cross-product statistic measured agreement between changes from separate windows. Each patient had equal weight. The mean near-minus-far score was 0.001579 squared-correlation units. Its descriptive 95 percent patient bootstrap interval was -0.001010 to 0.003965. This test did not establish greater local change or equivalence. The contact check outside both marked seizure zones changed sign, with an interval that also included zero. The native-rate whole-network result closely matched an earlier 256 Hz analysis of the same cohort. All 355 input files passed fresh hash and size checks, tied to the release Git commit. A full raw rebuild matched 490 compact arrays. A repeat matched all 43 numerical output files. The package supplies code, matrices, geometry, raw audit samples, edge and patient scores, three figures, and a full retrieval recipe. This exploratory result applies to offline voltage features in a reused epilepsy cohort. It does not measure a causal stimulation effect or BCI task accuracy.

stereoelectroencephalography · beta correlation · intracranial EEG · signal processing · spatial analysis · brain-computer interface · repeated measurements