An equal-weight interior Laplacian omitted a larger mean contrast variance fraction at 1–30 Hz than at 30–80 Hz in 36 SEEG patients

We tested an equal-weight Laplacian montage with no endpoint rows on public interictal SEEG data from 36 patients. The montage omitted mean contact contrast variance of 38.66% at 1–30 Hz and 33.43% at 30–80 Hz. The mean paired difference was 5.23 percentage points. Its descriptive 95% patient bootstrap interval was 3.07–7.37 points. The interval assumes independent, exchangeable patients. Each consecutive contact path lost one index-gradient contrast. One endpoint bipolar row per path restored the full contrast space. The score measures variance outside the montage row space. It does not measure clinical benefit or the limit of statistical prediction. This is a new secondary analysis of a reused cohort. All 72 fresh public EDF excerpts reproduced the input covariance exactly. The package preserves code, covariance, source versions, checks, bootstrap draws, and repeat instructions.

stereoelectroencephalography · SEEG · Laplacian montage · contact contrast variance · signal processing · OpenNeuro