Filter response correction reduced the forward and backward 60 Hz notch slope difference from unfiltered SEEG by 99.6% in 24 participants at 20–100 Hz with Q=30 and 55–65 Hz excluded

A forward and backward 60 Hz Q=30 notch changed a straight SEEG spectral slope by a mean absolute amount of 0.005219. We tested public signals from 24 participants in OpenNeuro ds005624, version 1.0.0. The fit used 20 to 100 Hz and excluded 55 to 65 Hz. Correction from the known digital filter response reduced the difference from the same unfiltered signal to 0.00001924. The reduction was 99.63 percent, with a 95 percent participant bootstrap interval of 99.62 to 99.64 percent. All 24 participants had a lower difference. Fixed checks used a wider notch, another fit band, a longer edge guard, and all adjacent pairs. Exact stationary sine removal left the fitted slope unchanged within numerical precision. The correction preserves within-record pair order and time differences. This result measures a small processing offset. It does not establish a true neural slope. The package preserves spectra, signal fixtures, source pins, code, results, numerical checks, and repeat instructions.

SEEG · stereoelectroencephalography · spectral slope · notch filter · 60 Hz · filter response correction · line noise · signal processing