In 171 SEEG patients, location-adjusted 4–13 Hz peak frequency correlated across hemispheres (r = 0.44, 95% CI 0.26–0.62; RAM FR1/catFR1, one 60 s task segment)

We tested if the 4-13 Hz spectral peak frequency of SEEG bipolar pairs is shared by distant regions of one patient. We used archived 0.5 Hz spectra of 27974 pairs in 256 patients (OpenNeuro ds004789 and ds004809, via AutoSci c3188d78). For each pair, we subtracted the mean peak of other patients' pairs within 15 mm (MNI). Then we averaged the adjusted peak per hemisphere. In 171 patients with pairs in both hemispheres, the left and right values correlated after site adjustment: r = 0.44 (95 % CI 0.26-0.62, permutation p = 0.0001; plan hashed before the analysis). Anterior vs posterior (r = 0.46, 205 patients) and temporal vs extra-temporal pairs (r = 0.44, 207 patients) gave the same result. Without location adjustment, r was 0.51. Split-half reliability inside each hemisphere was 0.74-0.81, and the disattenuated left-right r was 0.70 (CI 0.38-1.00). The ICC of single adjusted pair peaks was 0.16. Thus the SEEG peak frequency has a patient-level component (SD about 1 Hz) that distant regions share, and peak-frequency maps must model a patient offset. The causes (state, drugs, disease, trait) are not known.

stereo-EEG · SEEG · peak frequency · patient-level dependence · hemispheric correlation · location adjustment · memory task