80 to 150 Hz power concentration gave a mean local ranking fraction of 0.759 across four consecutive contact numbers in 30 SEEG patients

We tested 80 to 150 Hz power concentration across contact label transitions in public SEEG. The data came from OpenNeuro ds004100 version 1.1.3. Thirty patients each supplied two 60-second interictal clips. We used 78 comparisons between onset pairs and other pairs in the same contact name group. The pairs used four consecutive contact numbers and shared no contact. The mean local ranking fraction was 0.759, with a 95% patient bootstrap interval of 0.666 to 0.842. The broad comparison kept the same onset pairs and used all eligible other pairs in their groups. Its mean was 0.705. The planned local-minus-broad difference was 0.054, with interval -0.004 to 0.111. The test did not establish a change. A later exploratory score based only on minus contact number gave a local fraction of 0.813. Concentration had no clear advantage over that score. Source coordinate units conflict with the numeric scale, so local means contact-number proximity only. Fresh source retrieval reproduced all 68 feature files exactly. The package preserves code, source labels and identities, input powers, comparisons, patient values, bootstrap draws, figures, plans, checks, and reproduction steps. These results assume independent patients for the intervals. They do not establish clinical performance or metric anatomical proximity.

stereoelectroencephalography · epilepsy · seizure-onset labels · high-band power · ranking fraction · contact topology · interictal recordings