Delay frontal–temporal theta consistency tracked full recall, but did not rise above fixation, in 10 SEEG patients

We analyzed preprocessed bipolar SEEG from 13 patients performing a load-3 object working-memory task (OpenNeuro ds006136 v1.0.1). We measured theta (4–8 Hz) pairwise phase consistency (PPC) between middle frontal and middle temporal gyrus contacts. The pre-analysis plan specified two primary tests. Delay PPC was not shown to exceed fixation (mean difference +0.0016; 95% CI −0.0002 to +0.0036; 6/11 subjects positive; exact sign-flip p = 0.17, Holm p = 0.17). Delay PPC was higher on full-recall (3/3 objects) than partial-recall (0–2) trials (mean difference +0.0057; 95% CI +0.0019 to +0.0102; 8/10 subjects positive; exact p = 0.016, Holm p = 0.031). Fixation showed no recall difference (−0.0087, p = 0.79), and theta power did not differ between recall classes (p = 0.67). The effect is small and based on 10 subjects. A post hoc power-adjusted test was null across all subjects (p = 0.99); without one subject contributing five pairs, it was positive (p = 0.020). PPC includes zero-lag phase, so volume conduction may contribute. The data are preprocessed by the dataset authors, all participants had epilepsy, and secondary tests were not corrected for multiple comparisons. The package preserves the analysis plan, code, per-pair values, results, file URLs, and SHA-256 hashes.

SEEG · working memory · theta phase consistency · pairwise phase consistency · recall