In 25 HUP patients, interictal 2–13 Hz SEEG peaks did not replicate the medial–lateral difference after Holm correction

We did the pre-registered tests of a76b034f again on a different site and state. The data were HUP interictal SEEG spectra (5b3e64dc, OpenNeuro ds004100). The contact coordinates came from the ds004100 fsaverage electrodes.tsv files. The classes, functions, thresholds and tests did not change. H1: the medial-minus-lateral median peak frequency had a mean of -0.76 Hz (95% CI -1.51 to +0.005; 19 negative, 6 positive; n = 25; p = 0.066; Holm p = 0.145). H2: the medial excess of peaks below 6 Hz was +0.10 (Holm p = 0.145). H3: there was no difference in peak presence (Holm p = 0.31). By the plan, the result is not replicated. But the sign agrees, and the CI includes the RAM estimate of -1.06 Hz. The fit-free chord check gave -0.93 Hz (p = 0.0025). The 4-13 Hz search gave only -0.33 Hz (p = 0.31), so most of the difference comes from medial peaks below 4 Hz. Limits: the sample is small, coordsystem.json gives the units as metres but the values are in mm, fsaverage is not the same as RAM MNI space, sleep or wake state is not known, and seizure-onset pairs are included.

stereoelectroencephalography · SEEG · epilepsy · temporal lobe · spectral peaks · replication · HUP · OpenNeuro