The 10 Hz pre-stimulus magnitude effect did not replicate in 8 Mayo SEEG patients

A previous report found a small positive association between pre-stimulus 10 Hz wavelet magnitude at the response site and single-trial CCEP N1 amplitude in 8 UMC Utrecht patients (mean per-unit r = 0.055; all 8 positive; p = 0.0078). We tested this result with a pre-registered pipeline in an independent center using OpenNeuro ds004696 v1.0.1: 8 Mayo Clinic SEEG patients receiving 6 mA biphasic single pulses. Across 689 response-site units, the 10 Hz mean of patient mean correlations was r = -0.022 (2/8 positive; one-sided exact sign-flip p = 0.98). The pooled unit mean was r = -0.017 (95% bootstrap CI [-0.041, +0.007]), which excludes the earlier estimate of +0.055. Secondary 6 Hz, 15 Hz, and real-minus-sham tests were not significant after Holm correction. Sham windows 0.25 s before pulses produced effects of similar size; therefore, the unregistered two-sided 10 Hz result (p = 0.047) is not interpreted as a stimulation-specific negative effect. In this dataset, the 10 Hz pre-stimulus magnitude did not predict single-trial N1. The test included only 8 patients, and its patient-level sign-flip test has low power for effects near r = 0.05. Units in the pooled interval are treated as independent within patients. The centers also differed in stimulus intensity (6 mA versus 1–2 mA) and recording coverage (one shaft per patient here); the study did not test whether these differences explain the result.

cortico-cortical evoked potentials · CCEP · SEEG · single-trial N1 · pre-stimulus wavelet magnitude · replication · OpenNeuro ds004696 · intracranial stimulation