Sleep SEEG in 17 children: the HFA-envelope lag of bipolar pairs does not change along the anterior-posterior axis

Question: scalp EEG shows that NREM slow waves usually travel from anterior to posterior cortex. Do anterior SEEG depth pairs reach up and down states before posterior pairs? Method (pre-registered): we used 180 s of sleep SEEG from each of the 18 UCLA subjects in OpenNeuro ds005398. There were 1048 non-SOZ bipolar pairs with MNI coordinates. For each pair, we took the log 70-150 Hz envelope (0.2-4 Hz). This envelope does not depend on the slow-wave polarity. We found the lag of the pair envelope to the mean envelope of the other shafts (cross-correlation peak within +-300 ms). Result: in 17 patients, the mean Spearman rho(lag, y) was +0.002 (95% CI -0.064 to +0.070, sign-flip p = 0.95). The y slope with |x| and z in the model was -1.7 ms per 10 mm (CI -8.2 to +3.9). An event-locked lag gave +0.053 (Holm p = 0.27). A post hoc check showed that the two lag measures agreed within patients (mean rho +0.28), so the lags were not only noise. Limits: the sleep stage is not given, and the coupling to the common signal was weak (median peak r 0.08). The patients are children with epilepsy. There is one segment per patient. The S1 CI does not exclude a gradient of the scalp size (about -3.7 ms per 10 mm). The plan was changed before any signal was read because ds007118/9/20 SEEG contacts mostly have no coordinates.

sleep · SEEG · slow waves · high-frequency activity · epilepsy · anterior-posterior gradient