In a 7-patient pilot, hippocampal theta-phase/30–80 Hz amplitude coupling did not significantly increase with working-memory load (Holm p = 0.28)

This pilot tested whether hippocampal theta-phase to gamma-amplitude phase-amplitude coupling (PAC) during verbal working-memory maintenance increases with set size, as previously reported. The pre-specified analysis used session 1 from seven of 14 patients with hippocampal contacts in OpenNeuro ds004752 v1.0.1: 37 bipolar pairs and correct, artifact-free trials in a 3-s maintenance window. For the primary measure, Tort modulation index (18 phase bins) between 4–8 Hz phase and 30–80 Hz amplitude was z-scored against 200 trial-shuffle surrogates. Trial counts were equalized across set sizes 4, 6, and 8; pair slopes were averaged within patients and tested with an exact sign-flip test, with Holm correction across three amplitude bands. The primary slope was positive (+0.46 z per two items; six of seven patients), but was not significant (p = 0.094; Holm p = 0.28). Pooled-trial coupling was above zero in six of seven patients in each band (p = 0.031 per band; Holm p = 0.094). Secondary-band load slopes, raw and debiased MI slopes, and theta and gamma power load effects were not significant. The result is inconclusive: the pilot had seven patients, few trials per load, and one session. Trial-shuffle surrogates may be conservative if maintenance theta phase is aligned across trials; no interictal epileptiform discharge removal was done beyond the dataset artifact flag. The cohort was selected by data download success rather than data values. Full methods, per-subject and per-pair results, source hashes, code, and limitations are preserved in the accompanying text and original files.

phase-amplitude coupling · hippocampus · theta-gamma · verbal working memory · stereoelectroencephalography · OpenNeuro ds004752 · pilot study