Five second rank thresholds reduced mean beta duration excess over phase controls in nine selected adults with epilepsy

We tested beta duration in public rest and film SEEG from nine selected adults with epilepsy. The analysis used 513 adjacent bipolar pairs, a 13 to 25 Hz band, and 199 phase controls per block. Both rank rules selected exactly 20 percent of envelope samples. Long runs had at least 16 samples, or 250 ms. Whole-block ranks gave a mean excess of 4.098 percentage points over controls. Five second ranks reduced this excess by 1.641 points. The patient bootstrap 95 percent interval for reduction was 0.642 to 2.923 points. The Holm-adjusted two-sided sign-flip probability was 0.023438. The test assumes independent patients and symmetric null effects. The remaining local excess was 2.457 points. A fixed signal example showed a similar rule effect after a known slow gain change. The result shows dependence on threshold time interval. It does not identify a neural cause or detector accuracy. The reused clinical cohort makes the study exploratory. The package preserves compact inputs, source hashes, raw retrieval code, analysis code, full primary control scores, and exact repeated-run checks.

beta bursts · SEEG · duration score · Fourier phase controls · rank threshold · epilepsy · signal processing