Hippocampus–scalp theta and alpha phase locking did not increase with working-memory load in 13 patients

Question: does phase locking between hippocampal SEEG and scalp EEG during working-memory maintenance increase with memory load (4, 6 or 8 letters), as Boran et al. 2019 reported? We used both sessions of 13 patients in OpenNeuro ds004752 who have scalp EEG and a hippocampal contact. Per patient, we calculated the Spearman correlation between set size and the within-trial phase locking value (PLV) for each trial (3.0–5.75 s, theta 4–8 Hz and alpha 8–12 Hz). We tested the group with an exact sign-flip test and Holm correction. Response time increased with set size (12 of 13 patients, p = 0.0005), so the load manipulation is valid. The primary tests were null: theta mean Fisher z = −0.011 (p = 0.71, Holm p = 1.0), alpha +0.004 (p = 0.91, Holm p = 1.0). wPLI, power-residualised PLV and fixation-period PLV were also null. Across-trial theta PPC (set size 8 minus 4) and theta PLV to parietal channels (8 patients) had uncorrected p of about 0.02 with very small effects, but they did not survive Holm correction (p = 0.22). Limits: only 1–2 bipolar pairs on one hippocampal shaft per patient; recorded scalp reference; 6 patients have only 6 scalp channels; this is not an exact copy of the original pipeline. The package contains the plan, the numpy-only code, per-session source hashes and all results.

working memory · hippocampus · scalp EEG · SEEG · phase locking · phase-locking value · theta · alpha · verbal Sternberg task · OpenNeuro ds004752 · replication