A contact number rule gave more correct ranks than a fixed power score with the stored reference for 64 adjacent SEEG pairs with opposite onset marks in 19 patients

This study tests local SEEG ordering in public OpenNeuro ds003876 version 1.0.2. It compares a fixed low frequency power score with a fixed contact number rule. The sample has one 120-second interictal file from each of 19 patients at NIH and Johns Hopkins. The test uses 64 adjacent contact pairs with opposite supplied onset marks. Each patient has equal weight. Contact number gives 78.3 percent correct ordering. Stored-reference power gives 53.6 percent. The paired number advantage is 24.7 percentage points, with a 95 percent patient bootstrap interval of 7.9 to 42.3 points. The exact two-sided patient sign-flip p value is 0.0137. That test assumes independent patients and symmetric differences under the null. This extends an earlier HUP result to this external sample. A common average reference gives a smaller advantage with an interval that includes zero. Contact numbers do not prove physical depth or a tissue cause. The result does not establish clinical benefit. The package contains all source features, metadata, pair tables, code, resampling arrays, figures, raw retrieval recipes, and verification records.

SEEG · stereoelectroencephalography · seizure onset zone · interictal EEG · contact ordering · OpenNeuro · signal power